Tell us about your PDF experience.

Connect to data in Power BI -

documentation
Power BI documentation provides expert information for connecting to data with tools
such as gateways, template apps, and data refresh.

Connect to data in Power BI

ｐ CONCEPT

Data sources

Data view in Power BI Desktop

New name for Power BI datasets

ｇ TUTORIAL

Import data from a web page

Gateways

ｐ CONCEPT

On-premises data gateways

ｇ TUTORIAL

Connect to on-premises SQL Server data

ｃ HOW-TO GUIDE

Use personal gateways

Template apps

ｅ OVERVIEW

What are template apps?

ｐ CONCEPT

Create template apps

Distribute template apps in your org

Refresh data

ｐ CONCEPT

Data refresh

Incremental refresh for semantic models

ｃ HOW-TO GUIDE

Configure scheduled refresh

Data sources for the Power BI service
Article • 01/30/2024

Data is the core of Power BI. You can explore data by creating charts and dashboards or
by asking questions with Q&A. The visualizations and answers get their underlying data
from a semantic model, which comes from a data source.

This article focuses on data source types that you can connect to from the Power BI
service. There are many other types of data sources. To use these other data sources in
the Power BI service, you might first need to use Power BI Desktop or the advanced data
query and modeling features in Excel. For more information, see Databases and Other
data sources.

Discover content
You can use the OneLake data hub to discover existing data and reports.

On your Power BI site, select OneLake data hub in the navigation pane:

The tiles at the top of the page show recommended data items. For example, data can
be recommended because it's promoted by someone in your organization or because it
was accessed recently.
Below those tiles is a list of data that you have access to. You can filter to show all data,
your own data, or data endorsed by someone in your organization:

You can select Apps on the navigation pane to discover apps published by other people
in your organization. At the top right of that tab, select Get apps to choose apps from
online services that you use:
Many services have template apps for Power BI. Most services require an account. For
more information, see Connect to services you use with Power BI.

Create content
To create content, you can import or create files or databases.

Files
To import files:

1. Go to the workspace to which you want to import the files. Select New and then
Semantic model:

2. Select Excel or CSV. You can also paste or manually enter data.
When you import Excel or CSV files or manually create a workbook, Power BI imports
any supported data in tables and any data model into a new Power BI semantic model.

You can also upload files. Use this method for .pbix files. When you upload Excel files
from OneDrive or SharePoint, Power BI creates a connection to the file. When you
upload a local file, Power BI adds a copy of the file to the workspace.

To upload files, on the My workspace tab, select Upload to upload local files or files
from SharePoint or OneDrive:

Following are some types of files that you can add:

Excel workbooks, or .xlsx and .xlsm files, can include different data types. For
example, workbooks can include data that you enter into worksheets yourself, or
data that you query and load from external data sources by using Power Query.
Power Query is available via Get & Transform Data on the Data tab of Excel, or via
Get External Data in Power Pivot. You can import data from tables in worksheets or
import data from a data model. For more information, see Get data from files for
Power BI.
 Power BI Desktop, or .pbix report files, query and load data from external data
sources to create reports. In Power BI Desktop, you can extend your data model by
using measures and relationships, and publish the .pbix files to the Power BI
service. Power BI Desktop is intended for advanced users who have a thorough
understanding of their data sources, data querying and transformation, and data
modeling. For more information, see Connect to data in Power BI Desktop.

Comma-separated value, or .csv files, are simple text files with rows of data that
contain values separated by commas. For example, a .csv file that contains name
and address data might have many rows, each with values for first name, last
name, street address, city, and state. You can't import data into a .csv file, but many
applications, like Excel, can save simple table data as .csv files.

For other file types, like XML (.xml) or text (.txt), you can use Excel Get & Transform
Data to query, transform, and load the data first. You can then import the Excel file
into the Power BI service.

Where you store your files makes a significant difference. OneDrive provides the
greatest flexibility and integration with Power BI. You can also keep your files on your
local drive, but if you need to refresh the data, there are a few extra steps. For more
information, see Get data from files for Power BI.

Databases
You can connect Azure databases to Power BI to get analytics and reports that provide
real-time insights. For example, you can connect to Azure SQL Database and explore
data by creating reports in Power BI. Whenever you slice data or add a field to a
visualization, Power BI queries the database directly.

For more information, see:

Azure and Power BI

Azure SQL Database with DirectQuery
Azure Synapse Analytics with DirectQuery

You can also use Power BI Desktop or Excel to connect to, query, and load data into data
models for various other databases. You can then import the file into Power BI where a
semantic model exists. If you configure scheduled refresh, Power BI uses the
configuration and connection information from the file to connect directly to the data
source. Power BI queries for updates and loads the updates into the semantic model.
For more information, see Connect to data in Power BI Desktop.
Other data sources
You can use hundreds of different data sources with Power BI. The data must be in a
format consumable by the Power BI service. Power BI can then use the data to create
reports and dashboards and answer questions with Q&A.

Some data sources already have data formatted for the Power BI service. These sources
are similar to template apps from service providers like Google Analytics and Twilio. SQL
Server Analysis Services tabular model databases are also ready to use.

In other cases, you might need to query and load the data you want into a file. For
example, your organization might store logistics data in a data warehouse database on a
server. But the Power BI service can connect to that database and explore its data only if
it's a tabular model database. You can use Power BI Desktop or Excel to query and load
the logistics data into a tabular data model that you then save as a file. You can import
that file into Power BI where a semantic model exists.

If the logistics data in the database changes every day, you can refresh the Power BI
semantic model. When you import the data into the semantic model, you also import
the connection information from Power BI Desktop or the Excel file.

If you configure a scheduled refresh or do a manual refresh on the semantic model,

Power BI uses the connection information with other settings to connect directly to the
database. Power BI then queries for updates and loads those updates into the semantic
model. You probably need an on-premises data gateway to help secure any data
transfer between an on-premises server and Power BI. When the transfer is complete,
visualizations in reports and dashboards refresh automatically.

So even if you can't connect to your data source directly from the Power BI service, you
can still get your data into Power BI. It just takes a few more steps and maybe some help
from your IT department. For more information, see Data sources in Power BI Desktop.

Semantic models and data sources

You might see the terms semantic model and data source used synonymously. But
semantic models and data sources are two different things, although they're related.

Power BI creates a semantic model automatically when you connect to and import data
from a file, template app, or live data source. A semantic model contains information
about the data source and data source credentials. The semantic model also often
includes a subset of data copied from the data source. When you create visualizations in
reports and dashboards, you often look at data from the semantic model.
The data in a semantic model comes from a data source. For example, data could come
from the following data sources:

An online service like Google Analytics or QuickBooks

A database in the cloud like Azure SQL Database
A database or file on a local computer or a server in your organization

Data refresh
If you save your file on a local drive or a drive in your organization, you might need an
on-premises gateway to be able to refresh the semantic model in Power BI. The
computer that stores the file must be running during the refresh. You can also reimport
your file, or use Publish from Excel or Power BI Desktop, but those processes aren't
automated.

If you save your files on OneDrive for work or school or on a SharePoint team site, your
semantic model, reports, and dashboard are always up to date. Because both OneDrive
and Power BI are in the cloud, Power BI can connect directly to your files or import the
files into Power BI. Power BI connects about once every hour and checks for updates.
The semantic model and any visualizations refresh automatically if there are any
updates.

Template apps from services also automatically update, once a day in most cases. You
can manually refresh these apps, but whether you see updated data depends on the
service provider. Updates to template apps from people in your organization depend on
the data sources they use and how the app creator configured the refresh.

Azure databases like SQL Database, Azure Synapse Analytics, and Spark in Azure
HDInsight are cloud data sources. The Power BI service is also in the cloud, so Power BI
can connect to those data sources live by using DirectQuery. With DirectQuery, Power BI
is always in sync, and you don't need to set up a scheduled refresh.

SQL Server Analysis Services is a live connection to Power BI just like an Azure cloud
database. The difference is that the database is on a server in your organization. This
type of connection requires an on-premises gateway, which your IT department can
configure.

Data refresh is an important consideration when you use Power BI. For more
information, see Data refresh in Power BI.

Considerations and limitations

Data sources for the Power BI service have the following limitations. Other limitations
apply to specific features, but the following list applies to the full Power BI service:

Semantic model size limit. Semantic models stored in shared capacities in the
Power BI service have a 1-GB size limit. For larger semantic models, use Power BI
Premium.

Distinct values in a column. When a Power BI semantic model caches data in

Import mode, it can store a limit of 1,999,999,997 distinct values in a column.

Row limit. When you use DirectQuery, Power BI imposes a limit on the query
results that it sends to your underlying data source. If the query sent to the data
source returns more than one million rows, you see an error and the query fails.
The underlying data can still contain more than one million rows. You're unlikely to
reach this limit, because most reports aggregate the data into smaller sets of
results.

Column limit. The maximum number of columns allowed across all tables in a
semantic model is 16,000 columns. This limit applies to the Power BI service and to
semantic models Power BI Desktop uses. Power BI uses this limit to track the
number of both columns and tables in the semantic model, which means the
maximum number of columns is 16,000 minus one for each table in the semantic
model.

Data source user limit. The maximum number of data sources allowed per user is
1,000. This limit applies only to the Power BI service.

Single Sign On (SSO) considerations. DirectQuery models can enable SSO access
to their data sources, which allows the security in the source system to be implicitly
applied to the DAX queries executed by each user. SSO can be enabled for each
source connection, and might involve configuring a gateway or VNET for some
types of sources. You can read more about enabling SSO for gateways in the SSO
for data gateways article.

Querying the SSO-enabled DirectQuery model using a Service Principal (SPN) isn't
supported, since the SPN credential can't be passed through to the DirectQuery
source. Instead, use a User Principal (UPN) to execute such queries against the
SSO-enabled DirectQuery semantic model.

Related content
Connect to services you use with Power BI
Get data from files for Power BI
Data refresh in Power BI
DirectQuery in Power BI
What is an on-premises data gateway?
Data sources in Power BI Desktop
What is an on-premises data gateway?
Article • 05/28/2024

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

The on-premises data gateway acts as a bridge to provide quick and secure data
transfer between on-premises data (data that isn't in the cloud) and several Microsoft
cloud services. These cloud services include Power BI, PowerApps, Power Automate,
Azure Analysis Services, and Azure Logic Apps. By using a gateway, organizations can
keep databases and other data sources on their on-premises networks, yet securely use
that on-premises data in cloud services.

How the gateway works

For more information on how the gateway works, see On-premises data gateway
architecture.

Types of gateways
There are three different types of gateways, each for a different scenario:

On-premises data gateway: Allows multiple users to connect to multiple on-

premises data sources. With a single gateway installation, you can use an on-
premises data gateway with all supported services. This gateway is well-suited to
complex scenarios in which multiple people access multiple data sources.

On-premises data gateway (personal mode): Allows one user to connect to

sources and can’t be shared with others. An on-premises data gateway (personal
mode) can only be used with Power BI. This gateway is well-suited to scenarios in
which you’re the only person who creates reports, and you don't need to share any
data sources with others.

Virtual network data gateway: Allows multiple users to connect to multiple data
sources secured using virtual networks. No installation is required because it's a
Microsoft managed service. This gateway is well-suited to complex scenarios in
which multiple people access multiple data sources.

Use a gateway
There are five main steps for using a gateway:

1. Download and install the gateway on a local computer.

2. Configure the gateway based on your firewall and other network requirements.
3. Add gateway admins who can also manage and administer other network
requirements.
4. Use the gateway to refresh an on-premises data source.
5. Troubleshoot issues with the gateway.

Related content
Install the on-premises data gateway
Power BI implementation planning: Data gateways

More questions? Try the Power BI Community

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Discover data items in the OneLake data
hub
Article • 01/25/2024

The OneLake data hub makes it easy to find, explore, and use the Fabric data items in
your organization that you have access to. It provides information about the items and
entry points for working with them.

The data hub provides:

A filterable list of all the data items you can access

A gallery of recommended data items
A way of finding data items by workspace
A way to display only the data items of a selected domain
An options menu of things you can do with the data item

This article explains what you see on the data hub and describes how to use it.

Open the data hub

To open the data hub, select the OneLake data hub icon in the navigation pane.
Find items in the data items list
The data items list displays all the data items you have access to. To shorten the list, you
can filter by keyword or data-item type using the filters at the top of the list. If you
select the name of an item, you'll get to the item's details page. If you hover over an
item, you'll see three dots that open the options menu when you select them.

The list has four tabs to narrow down the list of data items.

ﾉ Expand table

Tab Description

All data Data items that you're allowed to find.

My data Data items that you own.

Endorsed in Endorsed data items in your organization that you're allowed to find. Certified
your org data items are listed first, followed by promoted data items. For more information
about endorsement, see the Endorsement overview

Favorites Data items that you've marked as favorites.

The columns of the list are described below.

 ﾉ Expand table

Column Description

Name The data item name. Select the name to open the item's details page.

Endorsement Endorsement status.

Owner Data item owner (listed in the All and Endorsed in your org tabs only).

Workspace The workspace the data item is located in.

Refreshed Last refresh time (rounded to hour, day, month, and year. See the details section
on the item's detail page for the exact time of the last refresh).

Next refresh The time of the next scheduled refresh (My data tab only).

Sensitivity Sensitivity, if set. Select the info icon to view the sensitivity label description.

Find items by workspace

Related data items are often grouped together in a workspace. To see the data items by
workspace, expand the Explorer pane and select the workspace you're interested in. The
data items you're allowed to see in that workspace will be displayed in the data items
list.
 ７ Note

The Explorer pane may list workspaces that you don't have access to if the
workspace contains items that you do have access to (through explicitly granted
permissions, for example). If you select such a workspace, only the items you have
access to will be displayed in the data items list.

Find recommended items

Use the tiles across the top of the data hub to find and explore recommended data
items. Recommended data items are data items that have been certified or promoted by
someone in your organization or have recently been refreshed or accessed. Each tile
contains information about the item and an options menu for doing things with the
item. When you select a recommended tile, you are taken to the item's details page.
Display only data items belonging to a
particular domain
If domains have been defined in your organization, you can use the domain selector to
select a domain so that only data items belonging to that domain will be displayed. If an
image has been associated with the domain, you’ll see that image on the data hub to
remind you of the domain you're viewing.

For more information about domains, see the Domains overview

Open an item's options menu

Each item shown in the data hub has an options menu that enables you to do things,
such as open the item's settings, manage item permissions, etc. The options available
depend on the item and your permissions on the item.

To display the options menu, select More options (...) on one of the items shown in the
data items list or a recommended item. In the data items list, you need to hover over the
item to reveal More options.
 ７ Note

The Explorer pane may list workspaces that you don't have access to if the
workspace contains items that you do have access to (through explicitly granted
permissions, for example). If you select such a workspace, only the items you have
access to will be displayed in the data items list.

Considerations and limitations

Streaming semantic models are not shown in the OneLake data hub.

Related content
Navigate to your items from Microsoft Fabric Home
Endorsement

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Semantic model details
Article • 11/10/2023

The semantic model details page helps you explore, monitor, and leverage semantic
models. When you select a semantic model in the data hub, a workspace, or other place
in Power BI, the details page for that semantic model opens.

The semantic model details page:

Shows you metadata about the semantic model, including description,

endorsement, and sensitivity.
Provides actions that you can perform on the semantic model, such as share,
refresh, create new, Analyze in Excel, and more.
Lists the reports and scorecards that are built on top of the semantic model.

The page header displays the semantic model name, endorsement (if any), and semantic
model owner. To send an email to the semantic model owner or the semantic model
certifier (if any), select the header and then select the name of the owner.

Supported actions
The semantic model details page enables you to perform a number of actions. The
actions available vary from user to user depending on their permissions on the data
item, and thus not all actions are available for all users.
Action Description On Action bar,
choose:

Download this Downloads the .pbix file for this semantic model. File > Download this
file file

Manage Opens the manage semantic model permissions page. File > Manage
permissions permissions

Settings Opens the semantic model settings page. File > Settings

Refresh now Launches a refresh of the semantic model. Refresh > Refresh
now

Schedule Opens the semantic model settings page where you Refresh > Schedule
refresh can set scheduled refresh. refresh

Share Opens the Share semantic model dialog. Share, or use the
Share this data tile.

Create a report Opens the report editing canvas where you can create Create a report >
from scratch a new report based on the semantic model. From scratch, or use
the Visualize this data
tile.

Create a report Creates a copy of the template in My Workspace. This Create a report >
from template action is only available if a related report template From template, or use
exists. the Visualize this data
tile.

Create a report Opens the formatted table editing canvas. Create a report > As
as formatted formatted table, or
table use the Visualize this
data tile.

Analyze in Launches Analyze in Excel using this semantic model. Analyze in Excel
Excel

Open lineage Opens the lineage view for the semantic model. Lineage > Open
view lineage view

Impact analysis Opens the impact analysis side pane for this semantic Lineage > Impact
model. analysis

Chat in Teams Invite people to start chatting in Teams. People you Chat in Teams
invite will receive a Teams chat message from you
with a link to this semantic model details page. If they
have access to the semantic model, the link will open
this semantic model details page in Teams.

Show tables Opens a side panel showing the semantic model's Show tables
tables. In the tables view you can create table
 Action Description On Action bar,
choose:

previews by selecting desired columns.

View semantic model metadata

The semantic model details section shows:

The name of the workspace where the item is located.

The exact time of the last refresh.
Endorsement status and certifier (if certified).
Sensitivity (if set).
Description (if any). You can create or edit the description from here.

Explore related reports

The explore related reports section shows you all the reports and scorecards that are
built on the semantic model. You can create a copy of an item by selecting the line the
item is on and selecting the Save a copy icon that appears. This section also shows you
usage metrics for the related items.

The columns in the list of related reports are:

 Name: Report name. If the name ends with (template), it means that this report has
been specially constructed to be used as a template. For example, "Sales
(template)".
Type: Item type, for example, report or scorecard.
Endorsement: Endorsement status.
Workspace: The name of the workspace where the related item is located.
Unique viewers: Shows the total number of unique users who viewed the item at
least once in the last 30 days, excluding the current day's views.
Views: Shows the total number of times an item was viewed in the last 30 days,
excluding the current day's views.

Visualize this data

To create a report based on the semantic model, select the Create report button on this
tile and choose the desired option.

Auto-create: Creates an auto-generated report from the semantic model.

From template: Creates a copy of the template in My workspace.
From scratch: Opens the report editing canvas to a new report built on the
semantic model. When you save your new report, it will be saved in the workspace
that contains the semantic model if you have write permissions on that workspace.
If you don't have write permissions on the workspace, or if you are a free user and
the semantic model resides in a Premium-capacity workspace, the new report will
be saved in My workspace.
As formatted table: Opens the formatted table editing canvas.

７ Note
 Only one template will be shown in the Create report drop-down, even if more than
one report template exists for this semantic model.

Share this data

You can share the semantic model with other users in your organization. Selecting the
Share semantic model button opens the Share semantic model dialog, where you can
choose which permissions to grant on the semantic model.

Data preview
Data preview enables you to view a selected table or columns from the semantic model.
You can also export the data to supported file formats or create a paginated report.

Prerequisites
The semantic model can be inside Premium or non-Premium workspaces. Classic
workspaces aren't supported. Read about new and classic workspaces.
You need Build permission for the semantic model.

Select data to preview

To preview a semantic models's data from the semantic model details page, select a
table or columns on the Tables side panel.
If you don't see the side panel, select Show tables on the action bar.

An entirely filled parent checkbox on the semantic model's table indicates that all its
sub-tables and columns have been selected. A partially filled parent checkbox means
that only a subset of them has been selected.
When you select a table or columns in a table, they will be displayed on the Table
preview page that opens.
Table preview may not show all of the data you've selected. To see more, you can export
or build a paginated report.

You can resize column widths using a drag handle next to the column headers. Resizing
columns can make the table preview more readable, especially for long column input
values.

Show query
Show query enables you to copy the DAX query used to create the table preview to the
clipboard. This makes it possible to reuse the query for future actions.

Back
At any time you can return to the semantic model details page by selecting the Back
button on the action bar. Selecting the Back button clears all your selections and brings
you back to semantic model details page.

７ Note

Table preview is intended to quickly explore the underlying data of tables within
your semantic model. You cannot view measures or select more than one table or
columns across tables. You can select Create paginated report for that.

Export data
Select the Export button on the Table preview page to export the data to one of the
supported file formats.

Build a paginated report

Select the Create paginated report button to open the editor.
 ７ Note

Data will change from underlying data to summarized data. You can switch to
underlying data using More options.

In the editor you can select multiple tables, measure, fields across tables, apply table
styles, change aggregates, and so on.

You can then export the report to any of the supported file formats, and the file will be
saved to your default downloads folder. Or you can save it as a paginated report to a
workspace of your choice. Paginated reports fully preserve your report formatting.

Switch from summarized to underlying data in the editor

Select More options (...) to switch from Summarized data to Underlying data.
Next steps
Use semantic models across workspaces
Create reports based on semantic models from different workspaces
Endorse your semantic model
Questions? Try asking the Power BI Community
Datamart details
Article • 11/10/2023

The datamart details page helps you explore, monitor, and use datamarts. When you
select a datamart in the data hub, the details page for that datamart opens.

The datamart details page:

Shows you metadata about the datamart, including description, endorsement, and
sensitivity, and connection string.
Provides actions that you can perform on the datamart, such as share, refresh,
create new, and Analyze in Excel.
Lists the reports that are built on top of the datamart.

Supported actions
The datamart details page enables you to perform actions on the datamart. The
following table lists all supported actions. On the datamart details page you may see
only a subset of this list, as only actions you have permission to perform are listed.

Action Description On Action bar, choose:

Manage Opens the manage datamart permissions page. File > Manage
permissions permissions

Settings Opens the datamart settings page. File > Settings

Refresh now Launches a refresh of the semantic model. Refresh > Refresh now

Schedule Opens the semantic model settings page where you Refresh > Schedule
refresh can set scheduled refresh. refresh
 Action Description On Action bar, choose:

Refresh history Opens Refresh history window where you see the Refresh > Refresh
time, duration, and status of each refresh. You can history
download the history as a .csv file.

Share Opens the Share datamart dialog. Sharing a Share, or use the Share
datamart allows recipients to build content based this data tile.
on the underlying semantic model and query the
corresponding SQL endpoint.

Create a report Opens the report editing canvas where you can Create a report > From
from scratch create a new report based on the datamart. scratch, or use the
Visualize this data tile.

Analyze in Launches Analyze in Excel using this datamart. Analyze in Excel

Excel

Open lineage Opens the lineage view for the datamart. Lineage > Open lineage
view view

Impact analysis Opens the impact analysis side pane for this Lineage > Impact
datamart. analysis

Edit Opens the datamart in the Datamart editor. Edit

View datamart metadata

The datamart details section shows:

The location of the datamart.

Endorsement status (and certifier, if certified).
The exact time of the last refresh.
Sensitivity, if set.
T-SQL connection string.
Description, if any.

Explore related reports

The See what already exists section shows you reports that are built on top of the
datamart's auto-generated semantic model. You can create a copy of a report by
selecting the line the item is on and clicking the Save a copy icon that appears. This
section also shows you usage metrics for the related items.

７ Note

Reports build on top of other semantic models created from the datamart aren't
shown in this section.

The columns in the list of related reports are:

Name: Report name. If the name ends with (template), the report has been
specially constructed to be used as a template. For example, "Sales (template)".
Type: Item type, for example, report or scorecard.
Location: The name of the workspace where the related item is located.
Refreshed: Time of last refresh.
Endorsement: Endorsement status.
Sensitivity: Sensitivity.
Unique viewers: Shows the total number of unique users who viewed the item at
least once in the last 30 days, excluding the current day's views.
Views: Shows the total number of times an item was viewed in the last 30 days,
excluding the current day's views.

Visualize this data

To create a report based on the semantic model, select the Create a report button on
this tile and choose the desired option.
 The Create from scratch option opens the report editing canvas to a new report
built on the semantic model.

When you save your new report, it's saved in the workspace that contains the
semantic model, if you have write permissions on that workspace.

If you don't have write permissions on the workspace, the report is saved in My
workspace. If you're a free user and the semantic model resides in a Premium-
capacity workspace, the report is saved in My workspace.

The Paginated report option opens the paginated report online editor. For
information about creating a paginated report using the online editor, see Create
exportable paginated reports in the Power BI service.

Share this data

You can share the datamart with other users in your organization. Selecting the Share
datamart button opens the Share datamart dialog. People you share the datamart with
can build content based on the underlying semantic model and query the
corresponding SQL endpoint.
Next steps
Create exportable paginated reports in the Power BI service
Endorse your content
Questions? Try asking the Power BI Community
Power BI for Microsoft Azure users
Do you work with data, manage infrastructure, or build applications in Microsoft Azure?
Do you want to get value from your data or applications by using Power BI? These
resources will help you get up to speed. Welcome!

Are you more of a Power BI consumer? Welcome to you, too. We suggest starting with
Power BI for consumers.

Get started creating with Power BI

Start with Power BI Desktop
Start with the Power BI service
What is Power BI Report Server?

Analyze your Azure SQL data

Connect to data in Azure SQL databases
Connect to data in Azure Synapse Analytics

Analyze data from other Azure services

Connect to Azure Stream Analytics
Visualize data from Azure Cosmos DB
Visualize data from Azure Data Explorer

Analyze your Azure costs and usage

$
Connect to Azure Consumption Insights

Embed Power BI in your own applications

Power BI embedded analytics
Tutorial: Embed Power BI content using a sample embed-for-your-
customers application

Enrich Power BI with Azure Machine Learning

Azure Machine Learning and Power BI
Quickstart: Connect to data in Power BI
Desktop
Article • 01/23/2024

In this quickstart, you connect to data using Power BI Desktop, which is the first step in
building data models and creating reports.

If you're not signed up for Power BI, sign up for a free trial before you begin.

Prerequisites
To complete the steps in this article, you need the following resources:

Download and install Power BI Desktop, which is a free application that runs on
your local computer. You can download Power BI Desktop directly, or you can
get it from the Microsoft Store .
Download this sample Excel workbook , and create a folder called C:\PBID-qs
where you can store the Excel file. Later steps in this quickstart assume that is the
file location for the downloaded Excel workbook.
For many data connectors in Power BI Desktop, Internet Explorer 10 (or newer) is
required for authentication.

Launch Power BI Desktop

Once you install Power BI Desktop, launch the application so it's running on your local
computer. You're presented with a Power BI tutorial. Follow the tutorial or close the
dialog to start with a blank canvas. The canvas is where you create visuals and reports
from your data.

Connect to data
With Power BI Desktop, you can connect to many different types of data. These sources
include basic data sources, such as a Microsoft Excel file. You can connect to online
services that contain all sorts of data, such as Salesforce, Microsoft Dynamics, Azure Blob
Storage, and many more.

To connect to data, from the Home ribbon select Get data.

The Get Data window appears. You can choose from the many different data sources to
which Power BI Desktop can connect. In this quickstart, use the Excel workbook that you
downloaded in Prerequisites.

Since this data source is an Excel file, select Excel from the Get Data window, then select
the Connect button.

Power BI prompts you to provide the location of the Excel file to which to connect. The
downloaded file is called Financial Sample. Select that file, and then select Open.
Power BI Desktop then loads the workbook and reads its contents, and shows you the
available data in the file using the Navigator window. In that window, you can choose
which data you would like to load into Power BI Desktop. Select the tables by marking
the checkboxes beside each table you want to import. Import both available tables.


Once you've made your selections, select Load to import the data into Power BI
Desktop.

View data in the Fields pane

Once you've loaded the tables, the Fields pane shows you the data. You can expand
each table by selecting the arrow beside its name. In the following image, the financials
table is expanded, showing each of its fields.

And that's it! You've connected to data in Power BI Desktop, loaded that data, and now
you can see all the available fields within those tables.

Related content
There are all sorts of things you can do with Power BI Desktop once you've connected to
data. You can create visuals and reports. Take a look at the following resource to get you
going:

Get started with Power BI Desktop

Tutorial: Shape and combine data in
Power BI Desktop
Article • 11/10/2023

With Power BI Desktop, you can connect to many different types of data sources, then
shape the data to meet your needs, enabling you to create visual reports to share with
others. Shaping data means transforming the data: renaming columns or tables,
changing text to numbers, removing rows, setting the first row as headers, and so on.
Combining data means connecting to two or more data sources, shaping them as
needed, then consolidating them into a single query.

In this tutorial, you'll learn how to:

Shape data by using Power Query Editor.

Connect to different data sources.
Combine those data sources, and create a data model to use in reports.

This tutorial demonstrates how to shape a query by using Power BI Desktop,

highlighting the most common tasks. The query used here is described in more detail,
including how to create the query from scratch, in Getting Started with Power BI
Desktop.

Power Query Editor in Power BI Desktop uses the right-click menus, and the Transform
ribbon. Most of what you can select in the ribbon is also available by right-clicking an
item, such as a column, and choosing from the menu that appears.

Shape data
To shape data in Power Query Editor, you provide step-by-step instructions for Power
Query Editor to adjust the data as it loads and presents the data. The original data
source isn't affected; only this particular view of the data is adjusted, or shaped.

The steps you specify (such as rename a table, transform a data type, or delete a
column) are recorded by Power Query Editor. Each time this query connects to the data
source, Power Query Editor carries out those steps so that the data is always shaped the
way you specify. This process occurs whenever you use Power Query Editor, or for
anyone who uses your shared query, such as on the Power BI service. Those steps are
captured, sequentially, in the Query Settings pane, under APPLIED STEPS. We’ll go
through each of those steps in this article.
1. Import the data from a web source. Select the Get data dropdown, then choose
Web.
2. Paste this URL into the From Web dialog and select OK.

https://www.fool.com/research/best-states-to-retire
3. In the Navigator dialog, select Table 1 , then choose Transform Data.

 Tip

Some information in the tables from the previous URL may change or be updated
occasionally. As a result, you may need to adjust the selections or steps in this
article accordingly.

1. The Power Query Editor window opens. You can see the default steps applied so
far, in the Query Settings pane under APPLIED STEPS.

Source: Connecting to the website.

Extracted Table from Html: Selecting the table.
Promoted Headers: Changing the top row of data into column headers.
Changed Type: Changing the column types, which are imported as text, to
their inferred types.
 

2. Change the table name from the default Table 1 to Retirement Data , then press
Enter.

3. The existing data is ordered by a weighted score, as described on the source web
page under Methodology . Let's add a custom column to calculate a different
score. We'll then sort the table on this column to compare the custom score's
ranking to the existing Rank.

4. From the Add Column ribbon, select Custom Column.

5. In the Custom Column dialog, in New column name, enter New score. For the
Custom column formula, enter the following data:

( [Quality of life] + [Housing cost] + [Healthcare cost and quality] +

[Crime rate rate] + [#"Public health/COVID-19 response"] + [Sales
taxes] + [#"Non-housing costs"] + [Weather] ) / 8

6. Make sure the status message is No syntax errors have been detected, and select
OK.

7. In Query Settings, the APPLIED STEPS list now shows the new Added Custom step
we just defined.
Adjust the data
Before we work with this query, let's make a few changes to adjust its data:

Adjust the rankings by removing a column.

For example, assume Weather isn't a factor in our results. Removing this column
from the query doesn't affect the other data.

Fix any errors.

Because we removed a column, we need to adjust our calculations in the New

score column by changing its formula.

Sort the data.

Sort the data based on the New score column, and compare to the existing Rank
column.

Replace the data.

We'll highlight how to replace a specific value and how to insert an applied step.

These changes are described in the following steps.

1. To remove the Weather column, select the column, choose the Home tab from the
ribbon, and then choose Remove Columns.
 ７ Note

The New score values haven't changed, due to the ordering of the steps.
Power Query Editor records the steps sequentially, yet independently, of each
other. To apply actions in a different sequence, you can move each applied
step up or down.

2. Right-click a step to see its context menu.

3. Move up the last step, Removed Columns, to just above the Added Custom step.
4. Select the Added Custom step.

Notice the New score column now shows Error rather than the calculated value.

There are several ways to get more information about each error. If you select the
cell without clicking on the word Error, Power Query Editor displays the error
information.
 If you select the word Error directly, Power Query Editor creates an Applied Step in
the Query Settings pane and displays information about the error. Because we
don't need to display error information anywhere else, select Cancel.

5. To fix the errors, there are two changes needed, removing the Weather column
name and changing the divisor from 8 to 7. You can make these changes in two
ways:

a. Right-click the Custom Column step and select Edit Settings. This brings up the
Custom Column dialog you used to create the New score column. Edit the
formula as described previously, until it looks like this:
b. Select the New score column, then display the column's data formula by
enabling the Formula Bar checkbox from the View tab.

Edit the formula as described previously, until it looks like this, then press Enter.

= Table.AddColumn(#"Removed Columns", "New score", each ( [Quality

of life] + [Housing cost] + [Healthcare cost and quality] + [Crime
rate rate] + [#"Public health/COVID-19 response"] + [Sales taxes] +
[#"Non-housing costs"] ) / 7)

Power Query Editor replaces the data with the revised values and the Added
Custom step completes with no errors.

７ Note

You can also select Remove Errors, by using the ribbon or the right-click
menu, which removes any rows that have errors. However, in this tutorial we
want to preserve all the data in the table.
6. Sort the data based on the New score column. First, select the last applied step,
Added Custom to display the most recent data. Then, select the drop-down
located next to the New score column header and choose Sort Descending.

The data is now sorted according to New score. You can select an applied step
anywhere in the list, and continue shaping the data at that point in the sequence.
Power Query Editor automatically inserts a new step directly after the currently
selected applied step.

7. In APPLIED STEPS, select the step preceding the custom column, which is the
Removed Columns step. Here we'll replace the value of the Housing cost ranking
in Oregon. Right-click the appropriate cell that contains Oregon's Housing cost
value, and then select Replace Values. Note which Applied Step is currently
selected.
 8. Select Insert.

Because we're inserting a step, Power Query Editor reminds us that subsequent
steps could cause the query to break.

9. Change the data value to 100.0.

Power Query Editor replaces the data for Oregon. When you create a new applied
step, Power Query Editor names it based on the action, in this case, Replaced
Value. If you have more than one step with the same name in your query, Power
Query Editor appends an increasing number to each subsequent applied step's
name.

10. Select the last Applied Step, Sorted Rows.

Notice the data has changed regarding Oregon's new ranking. This change occurs
because we inserted the Replaced Value step in the correct location, before the
Added Custom step.

We’ve now shaped our data to the extent we need to. Next let’s connect to
another data source, and combine data.
Combine data
The data about various states is interesting, and will be useful for building further
analysis efforts and queries. However, most data about states uses a two-letter
abbreviation for state codes, not the full name of the state. We need a way to associate
state names with their abbreviations.

There's another public data source that does provides that association, but it needs a
fair amount of shaping before we can connect it to our retirement table. To shape the
data, follow these steps:

1. From the Home ribbon in Power Query Editor, select New Source > Web.

2. Enter the address of the website for state abbreviations,

https://en.wikipedia.org/wiki/List_of_U.S._state_abbreviations , and then select
Connect.

The Navigator displays the content of the website.

3. Select Codes and abbreviations for U.S. states, federal district, territories, and
other regions.

 Tip
 It will take a bit of shaping to pare this table’s data down to what we want. Is
there a faster or easier way to accomplish the following steps? Yes, we could
create a relationship between the two tables, and shape the data based on
that relationship. The following example steps are helpful to learn for working
with tables. However, relationships can help you quickly use data from
multiple tables.

To get the data into shape, follow these steps:

1. Remove the top row. Because it's a result of the way that the web page’s table was
created, we don’t need it. From the Home ribbon, select Remove Rows > Remove
Top Rows.

The Remove Top Rows dialog appears. Specify 1 row to remove.

2. Promote the new top row to headers with Use First Row As Headers from the
Home tab, or from the Transform tab in the ribbon.

3. Because the Retirement Data table doesn't have information for Washington DC or
territories, we need to filter them from our list. Select the Name and status of
region_1 column's drop-down, then clear all checkboxes except State.
4. Remove all unneeded columns. Because we need only the mapping of each state
to its official two-letter abbreviation (Name and status of region and ANSI
columns), we can remove the other columns. First select the Name and status of
region column, then hold down the CTRL key and select the ANSI column. From
the Home tab on the ribbon, select Remove Columns > Remove Other Columns.

７ Note

The sequence of applied steps in Power Query Editor is important, and affects
how the data is shaped. It’s also important to consider how one step might
 impact another subsequent step. For example, if you remove a step from the
applied steps, subsequent steps might not behave as originally intended.

７ Note

When you resize the Power Query Editor window to make the width smaller,
some ribbon items are condensed to make the best use of visible space.
When you increase the width of the Power Query Editor window, the ribbon
items expand to make the most use of the increased ribbon area.

5. Rename the columns and the table. There are a few ways to rename a column: First
select the column, then either select Rename from the Transform tab on the
ribbon, or right-click and select Rename. The following image shows both options,
but you only need to choose one.
6. Rename the columns to State Name and State Code. To rename the table, enter the
Name State Codes in the Query Settings pane.
Combine queries
Now that we’ve shaped the State Codes table the way we want, let’s combine these two
tables, or queries, into one. Because the tables we now have are a result of the queries
we applied to the data, they’re often referred to as queries.

There are two primary ways of combining queries: merging and appending.

For one or more columns that you’d like to add to another query, you merge the
queries.
For one or more rows of data that you’d like to add to an existing query, you
append the query.

In this case, we want to merge the queries:

1. From the left pane of Power Query Editor, select the query into which you want the
other query to merge. In this case, it's Retirement Data.

2. Select Merge Queries > Merge Queries from the Home tab on the ribbon.
 You might be prompted to set the privacy levels, to ensure the data is combined
without including or transferring data you don't want transferred.

The Merge window appears. It prompts you to select which table you'd like
merged into the selected table, and the matching columns to use for the merge.

3. Select State from the Retirement Data table, then select the State Codes query.

When you select a matching columns, the OK button is enabled.

4. Select OK.

Power Query Editor creates a new column at the end of the query, which contains
the contents of the table (query) that was merged with the existing query. All
columns from the merged query are condensed into the column, but you can
Expand the table and include whichever columns you want.

5. To expand the merged table, and select which columns to include, select the
expand icon ( ).

The Expand window appears.

6. In this case, we want only the State Code column. Select that column, clear Use
original column name as prefix, and then select OK.

If we had left the checkbox selected for Use original column name as prefix, the
merged column would be named State Codes.State Code.

７ Note

If you want to explore how to bring in the State Codes table, you can
experiment a bit. If you don’t like the results, just delete that step from the
APPLIED STEPS list in the Query Settings pane, and your query returns to the
state prior to applying that Expand step. You can do this as many times as you
like until the expand process looks the way you want it.

We now have a single query (table) that combines two data sources, each of which
was shaped to meet our needs. This query can be a basis for interesting data
connections, such as housing cost statistics, quality of life, or crime rate in any
state.

7. To apply your changes and close Power Query Editor, select Close & Apply from
the Home ribbon tab.

The transformed semantic model appears in Power BI Desktop, ready to be used

for creating reports.
Next steps
For more information on Power BI Desktop and its capabilities, see the following
resources:

What is Power BI Desktop?

Query overview in Power BI Desktop
Data sources in Power BI Desktop
Connect to data in Power BI Desktop
Common query tasks in Power BI Desktop
Tutorial: Analyze webpage data by using
Power BI Desktop
Article • 01/12/2023

As a long-time soccer fan, you want to report on the UEFA European Championship
(Euro Cup) winners over the years. With Power BI Desktop, you can import this data
from a web page into a report and create visualizations that show the data. In this
tutorial, you learn how to use Power BI Desktop to:

Connect to a web data source and navigate across its available tables.
Shape and transform data in the Power Query Editor.
Name a query and import it into a Power BI Desktop report.
Create and customize a map and a pie chart visualization.

Connect to a web data source

You can get the UEFA winners data from the Results table on the UEFA European
Football Championship Wikipedia page at
https://en.wikipedia.org/wiki/UEFA_European_Football_Championship .

Web connections are only established using basic authentication. Web sites requiring
authentication might not work properly with the Web connector.

To import the data:

1. In the Power BI Desktop Home ribbon tab, drop down the arrow next to Get Data,
and then select Web.

７ Note

You can also select the Get Data item itself, or select Get Data from the Power
BI Desktop get started dialog, then select Web from the All or Other section
of the Get Data dialog, and then select Connect.

2. In the From Web dialog, paste the URL

https://en.wikipedia.org/wiki/UEFA_European_Football_Championship into the URL
text box, and then select OK.
 After you connect to the Wikipedia web page, the Navigator dialog shows a list of
available tables on the page. You can select any of the table names to preview its
data. The Results[edit] table has the data you want, although it's not exactly in the
shape you want. You'll reshape and clean up the data before loading it into your
report.

７ Note

The Preview pane shows the most recent table selected, but all selected
tables load into the Power Query Editor when you select Transform Data or
Load.

3. Select the Results[edit] table in the Navigator list, and then select Transform Data.
 A preview of the table opens in Power Query Editor, where you can apply
transformations to clean up the data.

Shape data in Power Query Editor

You want to make the data easier to scan by displaying only the years and the
countries/regions that won. You can use the Power Query Editor to perform these data
shaping and cleansing steps.

First, remove all the columns except for two from the table. Rename these columns as
Year and CountryRegion later in the process.

1. In the Power Query Editor grid, select the columns. Select Ctrl to select multiple
items.

2. Right-click and select Remove Other Columns, or select Remove Columns >
Remove Other Columns from the Manage Columns group in the Home ribbon
tab, to remove all other columns from the table.

or
This version of the imported data has the word Details appended to the year. You can
remove the extra word Details from the first column cells.

1. Select the first column.

2. Right-click, and select Replace Values, or select Replace Values from the
Transform group in the Home tab of the ribbon. This option is also found in the
Any Column group in the Transform tab.

or

3. In the Replace Values dialog, type Details in the Value To Find text box, leave the
Replace With text box empty, and then select OK to delete the word Details from
this column.
Some cells contain only the word "Year" rather than year values. You can filter the
column to only display rows that don't contain the word "Year".

1. Select the filter drop-down arrow on the column.

2. In the drop-down menu, scroll down and clear the checkbox next to the Year
option, and then select OK.

Since you're only looking at the final winners data now, you can rename the second
column to CountryRegion. To rename the column:
 1. Double-click or tap and hold in the second column header, or

Right-click the column header, and select Rename, or

Select the *column and select Rename from the Any Column group in the
Transform tab of the ribbon.

or

2. Type CountryRegion in the header and press Enter to rename the column.

You also want to filter out rows that have null values in the CountryRegion column.
You could use the filter menu as you did with the Year values, or you can:

1. Right-click on the CountryRegion cell in the 2020 row, which has the value null.

2. Select Text Filters > Does not Equal in the context menu to remove any rows that
contain that cell's value.
Import the query into Report View
Now that you've shaped the data the way you want, you're ready to name your query
"Euro Cup Winners" and import it into your report.

1. In the Query Settings pane, in the Name text box, enter Euro Cup Winners.

2. Select Close & Apply > Close & Apply from the Home tab of the ribbon.
The query loads into the Power BI Desktop Report view, where you can see it in the
Fields pane.

 Tip

You can always get back to the Power Query Editor to edit and refine your query by:

Selecting the More options ellipsis (...) next to Euro Cup Winners in the Fields
pane, and selecting Edit query, or
Selecting Transform data in the Queries group of the Home ribbon tab in
Report view.

Create a visualization
To create a visualization based on your data:

1. Select the CountryRegion field in the Fields pane, or drag it to the report canvas.
Power BI Desktop recognizes the data as country/region names, and automatically
creates a Map visualization.
 

2. Enlarge the map by dragging the handles in the corners so all the winning
country/region names are visible.

3. The map shows identical data points for every country/region that won a Euro Cup
tournament. To make the size of each data point reflect how often the
country/region has won, drag the Year field to Drag data fields here under Bubble
size in the lower part of the Visualizations pane. The field automatically changes to
a Count of Year measure, and the map visualization now shows larger data points
for countries/regions that have won more tournaments.
 

Customize the visualization

As you can see, it's very easy to create visualizations based on your data. It's also easy to
customize your visualizations to better present the data in ways that you want.

Format the map

You can change the appearance of a visualization by selecting it and then selecting the
Format (paint brush) icon in the Visualizations pane. For example, the "Germany" data
points in your visualization could be misleading, because West Germany won two
tournaments and Germany won one. The map superimposes the two points rather than
separating or adding them together. You can color these two points differently to
highlight this fact. You can also give the map a more descriptive and attractive title.

1. With the visualization selected, select the Format icon, and then select Visual >
Bubbles > Colors to expand the data color options.
2. Turn Show all to On, and then select the drop-down menu next to West Germany
and choose a yellow color.

3. Select General > Title to expand the title options, and in the Text field, type Euro
Cup Winners in place of the current title.
 4. Change Text color to red, size to 12, and Font to Segoe UI (Bold).

Your map visualization now looks like this example:

 

Change the visualization type

You can change the type of a visualization by selecting it and then selecting a different
icon at the top of the Visualizations pane. For example, your map visualization is
missing the data for the Soviet Union, because that country/region no longer exists on
the world map. Another type of visualization like a treemap or pie chart might be more
accurate, because it shows all the values.

To change the map to a pie chart, select the map and then choose the Pie chart icon in
the Visualizations pane.

 Tip
 You can use the Data colors formatting options to make "Germany" and
"West Germany" the same color.
To group the countries/regions with the most wins together on the pie chart,
select the ellipsis (...) at the upper right of the visualization, and then select
Sort by Count of Year.

Power BI Desktop provides a seamless end-to-end experience, from getting data from a
wide range of data sources and shaping it to meet your analysis needs, to visualizing
this data in rich and interactive ways. Once your report is ready, you can upload it to
Power BI and create dashboards based on it, which you can share with other Power BI
users.

See also
Microsoft Learn training for Power BI
Watch Power BI videos
Visit the Power BI Forum
Read the Power BI Blog
Tutorial: Analyze sales data from Excel
and an OData feed
Article • 11/10/2023

It's common to have data in multiple data sources. For example, you could have two
databases, one for product information, and another for sales information. With Power
BI Desktop, you can combine data from different sources to create interesting,
compelling data analyses and visualizations.

In this tutorial, you combine data from two data sources:

An Excel workbook with product information

An OData feed containing orders data

You're going to import each semantic model and do transformation and aggregation
operations. Then, you can use the two source's data to produce a sales analysis report
with interactive visualizations. Later, apply these techniques to SQL Server queries, CSV
files, and other data sources in Power BI Desktop.

７ Note

In Power BI Desktop, there are often a few ways to accomplish a task. For example,
you can right-click or use a More options menu on a column or cell to see more
ribbon selections. Several alternate methods are described in the following steps.

Import Excel product data

First, import product data from the Products.xlsx Excel workbook into Power BI Desktop.

1. Download the Products.xlsx Excel workbook and save it as Products.xlsx.

2. Select the arrow next to Get data in the Power BI Desktop ribbon's Home tab, and
then select Excel from the Common data sources menu.
 ７ Note

You can also select the Get data item itself, or select Get data from the Power
BI Get started dialog box, then select Excel or File > Excel in the Get Data
dialog box, and then select Connect.

3. In the Open dialog box, navigate to and select the Products.xlsx file, and then
select Open.

4. In the Navigator, select the Products table and then select Transform Data.
 A table preview opens in the Power Query Editor, where you can apply
transformations to clean up the data.

７ Note

You can also open the Power Query Editor by selecting Transform data from the
Home ribbon in Power BI Desktop, or by right-clicking or choosing More options
next to any query in the Report view, and selecting Transform data.

Clean up the columns

Your combined report uses the Excel workbook's ProductID, ProductName,
QuantityPerUnit, and UnitsInStock columns. You can remove the other columns.

1. In Power Query Editor, select the ProductID, ProductName, QuantityPerUnit, and

UnitsInStock columns. You can use Ctrl to select more than one column, or Shift to
 select columns next to each other.

2. Right-click any of the selected headers. Select Remove Other Columns from the
dropdown menu. You can also select Remove Columns > Remove Other Columns
from the Manage Columns group in the Home ribbon tab.

Import the OData feed's order data

Next, import the order data from the sample Northwind sales system OData feed.

1. In Power Query Editor, select New Source and then, from the Most Common
menu, select OData feed.
2. In the OData feed dialog box, paste the Northwind OData feed URL,
https://services.odata.org/V3/Northwind/Northwind.svc/ . Select OK.

3. In Navigator, select the Orders table, and then select OK to load the data into
Power Query Editor.
 ７ Note

In Navigator, you can select any table name, without selecting the checkbox,
to see a preview.

Expand the order data

You can use table references to build queries when connecting to data sources with
multiple tables, such as relational databases or the Northwind OData feed. The Orders
table contains references to several related tables. You can use the expand operation to
add the ProductID, UnitPrice, and Quantity columns from the related Order_Details
table into the subject (Orders) table.

1. Scroll to the right in the Orders table until you see the Order_Details column. It
contains references to another table and not data.
 2. Select the Expand icon ( ) in the Order_Details column header.

3. In the dropdown menu:

a. Select (Select All Columns) to clear all columns.

b. Select ProductID, UnitPrice, and Quantity, and then select OK.

After you expand the Order_Details table, three new nested table columns replace the
Order_Details column. There are new rows in the table for each order's added data.
Create a custom calculated column
Power Query Editor lets you create calculations and custom fields to enrich your data.
You can create a custom column that multiplies the unit price by item quantity to
calculate the total price for each order's line item.

1. In the Power Query Editor's Add Column ribbon tab, select Custom Column.

2. In the Custom Column dialog box, type LineTotal in the New column name field.

3. In the Custom column formula field after the =, enter [Order_Details.UnitPrice] *

[Order_Details.Quantity]. You can also select the field names from the Available
columns scroll box and select << Insert, instead of typing them.

4. Select OK.
 The new LineTotal field appears as the last column in the Orders table.

Set the new field's data type

When Power Query Editor connects to data, it makes a best guess as to each field's data
type for display purposes. A header icon indicates each field's assigned data type. You
can also look under Data Type in the Home ribbon tab's Transform group.

Your new LineTotal column has an Any data type, but it has currency values. To assign a
data type, right-click the LineTotal column header, select Change Type from the
dropdown menu, and then select Fixed decimal number.
 ７ Note

You can also select the LineTotal column, then select the arrow next to Data Type in
the Transform area of the Home ribbon tab, and then select Fixed decimal
number.

Clean up the orders columns

To make your model easier to work with in reports, you can delete, rename, and reorder
some columns.

Your report is going to use the following columns:

OrderDate
ShipCity
ShipCountry
Order_Details.ProductID
Order_Details.UnitPrice
Order_Details.Quantity
LineTotal

Select these columns and use Remove Other Columns as you did with the Excel data.
Or, you can select the non-listed columns, right-click on one of them, and select
Remove Columns.

You can rename the columns prefixed with "Order_Details." to make them easier to
read:

1. Double-click or tap and hold each column header, or right-click the column
header, and select Rename from the dropdown menu.

2. Delete the Order_Details. prefix from each name.

Finally, to make the LineTotal column easier to access, drag and drop it to the left, just
to the right of the ShipCountry column.
Review the query steps
Your Power Query Editor actions to shape and transform data are recorded. Each action
appears on the right in the Query Settings pane under APPLIED STEPS. You can step
back through the APPLIED STEPS to review your steps, and edit, delete, or rearrange
them if necessary. However, changing preceding steps is risky as that can break later
steps.

Select each of your queries in the Queries list on the left side of Power Query Editor, and
review the APPLIED STEPS in Query Settings. After you apply the previous data
transformations, the APPLIED STEPS for your two queries should look like this:

Products query
Orders query

 Tip

Underlying the applied steps are formulas written in the Power Query Language,
also known as the M language. To see and edit the formulas, select Advanced
Editor in the Query group of the Home tab of the ribbon.

Import the transformed queries

When you're satisfied with your transformed data and ready to import it into Power BI
Desktop Report view, select Close & Apply > Close & Apply in the Home ribbon tab's
Close group.

Once the data is loaded, the queries appear in the Fields list in the Power BI Desktop
Report view.
Manage the relationship between the semantic
models
Power BI Desktop doesn't require you to combine queries to report on them. However,
you can use the relationships between semantic models, based on common fields, to
extend, and enrich your reports. Power BI Desktop may detect relationships
automatically, or you can create them in the Power BI Desktop Manage Relationships
dialog box. For more information, see Create and manage relationships in Power BI
Desktop.

The shared ProductID field creates a relationship between this tutorial's Orders and
Products semantic models.

1. In Power BI Desktop Report view, select Manage relationships in the Modeling

ribbon tab's Relationships area.
2. In the Manage relationships dialog box, you can see that Power BI Desktop has
already detected and listed an active relationship between the Products and
Orders tables. To view the relationship, select Edit.

Edit relationship opens, showing details about the relationship.

 3. Power BI Desktop has auto-detected the relationship correctly, so you can select
Cancel and then Close.

In Power BI Desktop, on the left side, select Model to view and manage query
relationships. Double-click the arrow on the line connecting the two queries to open the
Edit relationship dialog and view or change the relationship.
To get back to Report view from Model view, select the Report icon.

Create visualizations using your data

You can create different visualizations in Power BI Desktop Review View to gain data
insights. Reports can have multiple pages, and each page can have multiple visuals. You
and others can interact with your visualizations to help analyze and understand data. For
more information, see Interact with a report in Editing view in Power BI service.

You can use both of your data sets, and the relationship between them, to help visualize
and analyze your sales data.
First, create a stacked column chart that uses fields from both queries to show the
quantity of each product ordered.

1. Select the Quantity field from Orders in the Fields pane at the right, or drag it
onto a blank space on the canvas. A stacked column chart is created showing the
total quantity of all products ordered.

2. To show the quantity of each product ordered, select ProductName from Products
in the Fields pane, or drag it onto the chart.

3. To sort the products by most to least ordered, select the More options ellipsis (...)
at the visualization's upper right, and then select Sort By > Quantity.

4. Use the handles at the corners of the chart to enlarge it so more product names
are visible.

Next, create a chart showing order dollar amounts (LineTotal) over time (OrderDate).

1. With nothing selected on the canvas, select LineTotal from Orders in the Fields
pane, or drag it to a blank space on the canvas. The stacked column chart shows
the total dollar amount of all orders.

2. Select the stacked chart, then select OrderDate from Orders, or drag it onto the
chart. The chart now shows line totals for each order date.

3. Drag the corners to resize the visualization and see more data.
  Tip

If you only see Years on the chart and only three data points, select the arrow
next to OrderDate in the Axis field of the Visualizations pane, and select
OrderDate instead of Date Hierarchy. Alternatively, you might need to select
Options and settings > Options from the File menu, and under Data Load,
clear the Auto date/time for new files option.

Finally, create a map visualization showing order amounts from each country or region.

1. With nothing selected on the canvas, select ShipCountry from Orders in the Fields
pane, or drag it to a blank space on the canvas. Power BI Desktop detects that the
data is country or region names. It then automatically creates a map visualization,
with a data point for each country or region with orders.

2. To make the data point sizes reflect each country's/region's order amounts, drag
the LineTotal field onto the map. You can also drag it to Add data fields here
under Size in the Visualizations pane. The sizes of the circles on the map now
reflect the dollar amounts of the orders from each country or region.
Interact with your report visuals to analyze
further
In Power BI Desktop, you can interact with visuals that cross-highlight and filter each
other to uncover further trends. For more information, see Filters and highlighting in
Power BI reports.

Because of the relationship between your queries, interactions with one visualization
affect all the other visualizations on the page.

On the map visualization, select the circle centered in Canada. The other two
visualizations filter to highlight the Canadian line totals and order quantities.

Select a Quantity by ProductName chart product to see the map and the date chart
filter to reflect that product's data. Select a LineTotal by OrderDate chart date to see the
map and the product chart filter to show that date's data.

 Tip

To clear a selection, select it again, or select one of the other visualizations.

Complete the sales analysis report

Your completed report combines data from the Products.xlsx Excel file and the
Northwind OData feed in visuals that help you analyze different countries' or regions'
order information, time frames, and products. When your report is ready, you can
upload it to the Power BI service to share it with other Power BI users.

Next steps
Microsoft Learn training for Power BI
Watch Power BI videos
Visit the Power BI Forum
Read the Power BI Blog
Implement row-level security in an on-
premises Analysis Services tabular
model
Article • 01/16/2024

Using a sample semantic model to work through the steps below, this tutorial shows
you how to implement row-level security in an on-premises Analysis Services Tabular
Model and use it in a Power BI report.

Create a new security table in the AdventureworksDW2012 database

Build the tabular model with necessary fact and dimension tables
Define user roles and permissions
Deploy the model to an Analysis Services tabular instance
Build a Power BI Desktop report that displays data tailored to the user accessing
the report
Deploy the report to Power BI service
Create a new dashboard based on the report
Share the dashboard with your coworkers

This tutorial requires the AdventureworksDW2012 database .

Task 1: Create the user security table and define

data relationship
You can find many articles describing how to define row-level dynamic security with the
SQL Server Analysis Services (SSAS) tabular model.

The steps here require using the AdventureworksDW2012 relational database.

1. In AdventureworksDW2012, create the DimUserSecurity table as shown below. You

can use SQL Server Management Studio (SSMS) to create the table.
2. Once you create and save the table, you need to establish the relationship
between the DimUserSecurity table's SalesTerritoryID column and the
DimSalesTerritory table's SalesTerritoryKey column, as shown below.

In SSMS, right-click DimUserSecurity, and select Design. Then select Table

Designer > Relationships.... When done, save the table.

3. Add users to the table. Right-click DimUserSecurity and select Edit Top 200 Rows.
Once you've added users, the DimUserSecurity table should appear similar to the
following example:

You'll see these users in upcoming tasks.

4. Next, do an inner join with the DimSalesTerritory table, which shows the user
associated region details. The SQL code here does the inner join, and the image
shows how the table then appears.

SQL

select b.SalesTerritoryCountry, b.SalesTerritoryRegion, a.EmployeeID,

a.FirstName, a.LastName, a.UserName from [dbo].[DimUserSecurity] as a
 join [dbo].[DimSalesTerritory] as b on a.[SalesTerritoryID] = b.
[SalesTerritoryKey]

The joined table shows who is responsible for each sales region, thanks to the
relationship created in Step 2. For example, you can see that Rita Santos is
responsible for Australia.

Task 2: Create the tabular model with facts and

dimension tables
Once your relational data warehouse is in place, you need to define the tabular model.
You can create the model using SQL Server Data Tools (SSDT). For more information, see
Create a New Tabular Model Project.

1. Import all the necessary tables into the model as shown below.

2. Once you've imported the necessary tables, you need to define a role called
SalesTerritoryUsers with Read permission. Select the Model menu in SQL Server
Data Tools, and then select Roles. In Role Manager, select New.
3. Under Members in the Role Manager, add the users that you defined in the
DimUserSecurity table in Task 1.

4. Next, add the proper functions for both DimSalesTerritory and DimUserSecurity
tables, as shown below under Row Filters tab.
5. The LOOKUPVALUE function returns values for a column in which the Windows user
name matches the one the USERNAME function returns. You can then restrict queries
to where the LOOKUPVALUE returned values match ones in the same or related table.
In the DAX Filter column, type the following formula:

DAX

=DimSalesTerritory[SalesTerritoryKey]=LOOKUPVALUE(DimUserSecurity[Sales
TerritoryID], DimUserSecurity[UserName], USERNAME(),
DimUserSecurity[SalesTerritoryID],
DimSalesTerritory[SalesTerritoryKey])

In this formula, the LOOKUPVALUE function returns all values for the
DimUserSecurity[SalesTerritoryID] column, where the DimUserSecurity[UserName]

is the same as the current logged on Windows user name, and

DimUserSecurity[SalesTerritoryID] is the same as the

DimSalesTerritory[SalesTerritoryKey] .

） Important
 When using row-level security, the DAX function USERELATIONSHIP is not
supported.

The set of Sales SalesTerritoryKey 's LOOKUPVALUE returns is then used to restrict
the rows shown in the DimSalesTerritory . Only rows where the SalesTerritoryKey
value is in the IDs that the LOOKUPVALUE function returns are displayed.

6. For the DimUserSecurity table, in the DAX Filter column, add the following
formula:

DAX

=FALSE()

This formula specifies that all columns resolve to false ; meaning DimUserSecurity
table columns can't be queried.

Now you need to process and deploy the model. For more information, see Deploy.

Task 3: Add Data Sources within your On-

premises data gateway
Once your tabular model is deployed and ready for consumption, you need to add a
data source connection to your on-premises Analysis Services tabular server.

1. To allow the Power BI service access to your on-premises analysis service, you need
an on-premises data gateway installed and configured in your environment.

2. Once the gateway is correctly configured, you need to create a data source
connection for your Analysis Services tabular instance. For more information, see
Manage your data source - Analysis Services.
With this procedure complete, the gateway is configured and ready to interact with your
on-premises Analysis Services data source.

Task 4: Create report based on analysis services

tabular model using Power BI desktop
1. Start Power BI Desktop and select Get data > Database.

2. From the data sources list, select the SQL Server Analysis Services Database and
select Connect.
3. Fill in your Analysis Services tabular instance details and select Connect live. Then
select OK.

With Power BI, dynamic security works only with a live connection.
 4. You can see that the deployed model is in the Analysis Services instance. Select the
respective model and then select OK.

Power BI Desktop now displays all the available fields, to the right of the canvas in
the Fields pane.

5. In the Fields pane, select the SalesAmount measure from the FactInternetSales
table and the SalesTerritoryRegion dimension from the SalesTerritory table.

6. To keep this report simple, we won't add any more columns right now. To have a
more meaningful data representation, change the visualization to Donut chart.

7. Once your report is ready, you can directly publish it to the Power BI portal. From
the Home ribbon in Power BI Desktop, select Publish.

Task 5: Create and share a dashboard

You've created the report and published it to the Power BI service. Now you can use the
example created in previous steps to demonstrate the model security scenario.

In the role as Sales Manager, the user Grace can see data from all the different sales
regions. Grace creates this report and publishes it to the Power BI service. This report
was created in the previous tasks.

Once Grace publishes the report, the next step is to create a dashboard in the Power BI
service called TabularDynamicSec based on that report. In the following image, notice
that Grace can see the data corresponding to all the sales region.
Now Grace shares the dashboard with a colleague, Rita, who is responsible for the
Australia region sales.
When Rita logs in to the Power BI service and views the shared dashboard that Grace
created, only sales from the Australia region are visible.

Congratulations! The Power BI service shows the dynamic row-level security defined in
the on-premises Analysis Services tabular model. Power BI uses the EffectiveUserName
property to send the current Power BI user credential to the on-premises data source to
run the queries.
Task 6: Understand what happens behind the
scenes
This task assumes you're familiar with SQL Server Profiler, since you need to capture a
SQL Server profiler trace on your on-premises SSAS tabular instance.

The session gets initialized as soon as the user, Rita, accesses the dashboard in the
Power BI service. You can see that the salesterritoryusers role takes an immediate effect
with the effective user name as
<EffectiveUserName>[email protected]</EffectiveUserName>

<PropertyList><Catalog>DefinedSalesTabular</Catalog>
<Timeout>600</Timeout><Content>SchemaData</Content><Format>Tabular</Format>
<AxisFormat>TupleFormat</AxisFormat><BeginRange>-1</BeginRange>
<EndRange>-1</EndRange><ShowHiddenCubes>false</ShowHiddenCubes>
<VisualMode>0</VisualMode><DbpropMsmdFlattened2>true</DbpropMsmdFlattened2>
<SspropInitAppName>PowerBI</SspropInitAppName>
<SecuredCellValue>0</SecuredCellValue><ImpactAnalysis>false</ImpactAnalysis>
<SQLQueryMode>Calculated</SQLQueryMode>
<ClientProcessID>6408</ClientProcessID><Cube>Model</Cube>
<ReturnCellProperties>true</ReturnCellProperties>
<CommitTimeout>0</CommitTimeout><ForceCommitTimeout>0</ForceCommitTimeout>
<ExecutionMode>Execute</ExecutionMode><RealTimeOlap>false</RealTimeOlap>
<MdxMissingMemberMode>Default</MdxMissingMemberMode>
<DisablePrefetchFacts>false</DisablePrefetchFacts>
<UpdateIsolationLevel>2</UpdateIsolationLevel>
<DbpropMsmdOptimizeResponse>0</DbpropMsmdOptimizeResponse>
<ResponseEncoding>Default</ResponseEncoding>
<DirectQueryMode>Default</DirectQueryMode><DbpropMsmdActivityID>4ea2a372-
dd2f-4edd-a8ca-1b909b4165b5</DbpropMsmdActivityID>
<DbpropMsmdRequestID>2313cf77-b881-015d-e6da-
eda9846d42db</DbpropMsmdRequestID><LocaleIdentifier>1033</LocaleIdentifier>
<EffectiveUserName>[email protected]</EffectiveUserName></PropertyList>

Based on the effective user name request, Analysis Services converts the request to the
actual contoso\rita credential after querying the local Active Directory. Once Analysis
Services gets the credential, Analysis Services returns the data the user has permission
to view and access.

If more activity occurs with the dashboard, with SQL Profiler you would see a specific
query coming back to the Analysis Services tabular model as a DAX query. For example,
if Rita goes from the dashboard to the underlying report, the following query occurs.
You can also see below the DAX query that is getting executed to populate report data.

DAX

EVALUATE
ROW(
"SumEmployeeKey", CALCULATE(SUM(Employee[EmployeeKey]))
)

<PropertyList xmlns="urn:schemas-microsoft-com:xml-analysis">``
<Catalog>DefinedSalesTabular</Catalog>
<Cube>Model</Cube>
<SspropInitAppName>PowerBI</SspropInitAppName>
<EffectiveUserName>[email protected]</EffectiveUserName>
<LocaleIdentifier>1033</LocaleIdentifier>
<ClientProcessID>6408</ClientProcessID>
<Format>Tabular</Format>
<Content>SchemaData</Content>
<Timeout>600</Timeout>
<DbpropMsmdRequestID>8510d758-f07b-a025-8fb3-
a0540189ff79</DbpropMsmdRequestID>
<DbPropMsmdActivityID>f2dbe8a3-ef51-4d70-a879-
5f02a502b2c3</DbPropMsmdActivityID>
<ReturnCellProperties>true</ReturnCellProperties>
<DbpropMsmdFlattened2>true</DbpropMsmdFlattened2>
<DbpropMsmdActivityID>f2dbe8a3-ef51-4d70-a879-
5f02a502b2c3</DbpropMsmdActivityID>
</PropertyList>

Considerations
On-premises row-level security with Power BI is only available with live connection.

Any changes in the data after processing the model would be immediately
available for the users accessing the report with live connection from the Power BI
service.
Tutorial: Connect to a GitHub repo with
Power BI
Article • 11/10/2023

In this tutorial, you connect to real data: the Power BI content public repository (also
known as a repo) in the GitHub service. Power BI automatically creates a dashboard and
report with the data. You see answers to questions like: How many people contribute to
the Power BI public repo? Who contributes the most? Which day of the week has the
most contributions? And other questions.

You can connect to your own private or public GitHub repos too. To use a Power BI
template app to connect to your repos, see Connect to GitHub with Power BI.

In this tutorial, you complete the following steps:

＂ Sign up for a GitHub account, if you don't have one yet.

＂ Sign in to your Power BI account, or sign up, if you don't have one yet.
＂ Open the Power BI service.
＂ Find the GitHub app.
＂ Enter the information for the Power BI public GitHub repo.
＂ View the dashboard and report with GitHub data.
＂ Clean up resources by deleting the app.

If you're not signed up for Power BI, sign up for a free trial before you begin.

Prerequisites
To complete this tutorial, you need a GitHub account, if you don't already have one.

Sign up for a GitHub account.

How to connect
1. Sign in to the Power BI service (app.powerbi.com).

2. In the nav pane, select Apps, then Get apps.

3. Enter GitHub in the search box. Select the app, and then choose Get it now.

4. Select Install.
5. When you see the notification, Your new app is ready!, select Go to app.

6. On the app landing page, select Connect your data.

7. In the connect dialog, enter the repository name and repository owner of the repo.
The URL for this repo is https://github.com/MicrosoftDocs/powerbi-docs . Enter
MicrosoftDocs as the repository Owner, and powerbi-docs as the Repo name.

Select Next.
8. Make sure that Authentication Method is set to OAuth2 , and then select Sign in
and connect.
 9. If prompted, follow the GitHub authentication instructions and give Power BI
permission to access your data.

After Power BI can connect with GitHub, the data in your Power BI semantic
model is refreshed once a day.
After Power BI imports the data, you see the contents in your new GitHub
workspace.

10. Select Workspaces in the nav pane to see the dashboard, reports, and semantic
models. You can select More options (...) to view settings.
11. In workspace Settings, you can rename or delete the workspace.
12. Select your GitHub dashboard. You can minimize or expand the nav pane, so you
have more room to see your data.

The GitHub dashboard contains live data, so the values you see may be different.
Ask a question
1. Select the Ask a question about your data text box. Power BI opens the Q&A
window and offers some sample questions.

2. Enter how many users, Power BI offers a list of questions.

3. You can edit your question, for example, in between how many and users, type
pull requests per.

Power BI creates a bar chart visual that shows the number of pull requests per
person.
 4. Select the pin icon to pin the visual to your dashboard, then Exit Q&A.

View the GitHub report

1. On the GitHub dashboard, select More options (...) on the column chart Pull
Requests by Month. Choose Go to report.

2. Select a user name in the Total pull requests by user chart. A new tile appears with
results for one user.
 3. Select the Punch Card tab to view the next page of the report. Now you can see
volumes of work by hour of the day and day of the week.

Clean up resources
Now that you've finished the tutorial, you can delete the GitHub app.

1. In the nav pane, select Apps.

 2. On the app tile, select More options (...) and then choose Delete.

Next steps
In this tutorial, you've connected to a GitHub public repo and gotten data, which Power
BI has formatted in a dashboard and report. You've answered some questions about the
data by exploring the dashboard and report. Now you can learn more about connecting
to other services, such as Salesforce, Microsoft Dynamics, and Google Analytics.

Connect to GitHub with a Power BI template app

Tutorial: Use Cognitive Services in Power
BI
Article • 06/01/2023

Power BI provides access to a set of functions from Azure Cognitive Services to enrich
your data in the self-service data prep for dataflows. The services that are supported
today are sentiment analysis, key phrase extraction, language detection, and image
tagging. The transformations run on the Power BI service and don't require an Azure
Cognitive Services subscription. This feature requires Power BI Premium.

Cognitive Services transforms are supported in the self-service data prep for
dataflows . Use the step-by-step examples for text analytics and image tagging in this
article to get started.

In this tutorial, you learn how to:

＂ Import data into a dataflow

＂ Score sentiment and extract key phrases of a text column in a dataflow
＂ Connect to the results from Power BI Desktop

Prerequisites
To complete this tutorial, you need the following prerequisites:

A Power BI account. If you're not signed up for Power BI, sign up for a free trial
before you begin.
Access to a Power BI Premium capacity with the AI workload enabled. This
workload is turned off by default during preview. If you are in on a Premium
capacity and AI Insights aren't showing up, contact your Premium capacity
administrator to enable the AI workload in the admin portal.

Text analytics
Follow the steps in this section to complete the text analytics portion of the tutorial.

Step 1: Apply sentiment scoring in the Power BI service

To get started, navigate to a Power BI workspace with Premium capacity and create a
new dataflow using the Create button in the upper right of the screen.
The dataflow dialog shows you the options for creating a new dataflow, select Add new
entities. Next, choose Text/CSV from the menu of data sources.

Paste this URL into the URL field:

https://pbiaitutorials.blob.core.windows.net/textanalytics/FabrikamComments.csv

and select Next.

 

The data is now ready to use for text analytics. You can use Sentiment Scoring and Key
Phrase Extraction on the customer comments column.

In Power Query Editor, select AI Insights

Expand the Cognitive Services folder and select the function you would like to use. This
example scores the sentiment of the comment column, but you can follow the same
steps to try out Language Detection and Key Phrase Extraction.
 

After you select a function, the required and optional fields appear. To score the
sentiment of the example reviews, select the reviews column as text input. Culture
information is an optional input and requires an ISO format. For example, enter en if you
want the text to be treated as English. When the field is left blank, Power BI first detects
the language of the input value before it scores the sentiment.

Now select Invoke to run the function. The function adds a new column with the
sentiment score for each row to the table. You can go back to AI insights to extract key
phrases of the review text in the same way.

Once you finish the transformations, change the query name to Customer comments and
select Done.
 

Next, Save the dataflow and name it Fabrikam. Select the Refresh Now button that pops
up after you save the dataflow.

After you save and refresh the dataflow, you can use it in a Power BI report.

Step 2: Connect from Power BI Desktop

Open Power BI Desktop. In the Home ribbon, select Get data.

Select Power BI and then choose Power BI dataflows. Select Connect.

Sign in with your organization account.

Select the dataflow you created. Navigate to the Customer comments table and choose
Load.
 

Now that the data is loaded, you can start building a report.

Image tagging
In the Power BI service, navigate to a workspace with Premium capacity. Create a new
dataflow using the Create button in the upper right of the screen.

Select Add new entities.

Once you're asked to choose a data source, select Blank query.

Copy this query in the query editor and select Next. You can replace the URL paths with
other images or add more rows. The Web.Contents function imports the image URL as
binary. If you have a data source with images stored as binary, you can also use that
directly.

Python

let
Source = Table.FromRows({
{ Web.Contents("https://images.pexels.com/photos/87452/flowers-background-
butterflies-beautiful-87452.jpeg") },
{
Web.Contents("https://upload.wikimedia.org/wikipedia/commons/5/53/Colosseum_
in_Rome%2C_Italy_-_April_2007.jpg") }}, { "Image" })
 in
Source

When prompted for credentials, select anonymous.

You see the following dialog.

Power BI prompts you for credentials for each web page.

Select AI Insights in the query editor.

Next, sign in with your organizational account.

Select the Tag Images function, enter [Binary] in the column field, and enter en in the
culture info field.

７ Note

You currently cannot pick a column using a dropdown. This issue will be resolved as
soon as possible during the private preview.

In the function editor, remove the quotation marks around the column name.
 ７ Note

Removing the quotation marks is a temporary workaround. This issue will be

resolved as soon as possible during preview.

The function returns a record with both the tags in comma-separated format and as a
json record. Select the expand-button to add one or both as columns to the table.

Select Done and save the dataflow. Once you've refreshed the dataflow one, you can
connect to it from Power BI Desktop using the Dataflows connectors.

Clean up resources
When you're done using this tutorial, delete the query by right-clicking the query name
in the Power Query Editor and selecting Delete.

Limitations
There are some known issues with using Gateway with Cognitive Services. If you need to
use a gateway, we recommend creating a dataflow that imports the necessary data by
using a gateway first. Then create another dataflow that references the first dataflow to
apply these functions.
If your AI work with dataflows fails, you may need to enable Fast Combine when using
AI with dataflows. Once you have imported your table and before you begin to add AI
features, select Options from the Home ribbon, and in the window that appears select
the checkbox beside Allow combining data from multiple sources to enable the feature,
then select OK to save your selection. Then you can add AI features to your dataflow.

Next steps
In this tutorial, you applied sentiment scoring and image tagging functions on a Power
BI dataflow. To learn more about Cognitive Services in Power BI, see the following
articles.

Azure Cognitive Services

Get started with self-service data prep on dataflows
Learn more about Power BI Premium

You might also be interested in the following articles.

Tutorial: Consume Azure Machine Learning models in Power BI

AI with dataflows
Tutorial: Build a machine learning model
in Power BI
Article • 04/25/2024

） Important

Creation of Power BI Automated Machine Learning (AutoML) models for dataflows

v1 has been retired, and is no longer available. Customers are encouraged to
migrate your solution to the AutoML feature in Microsoft Fabric. For more
information, see the retirement announcement .

In this tutorial, you use automated machine learning to create and apply a binary
prediction model in Power BI. You create a Power BI dataflow, and use the entities you
define in the dataflow to train and validate a machine learning model directly in Power
BI. You then use that model to score new data and generate predictions.

First, you create a binary prediction machine learning model to predict the purchase
intent of online shoppers, based on a set of their online session attributes. You use a
benchmark machine learning semantic model for this exercise. Once you train a model,
Power BI automatically generates a validation report that explains the model results. You
can then review the validation report and apply the model to your data for scoring.

This tutorial consists of the following steps:

＂ Create a dataflow with the input data.

＂ Create and train a machine learning model.
＂ Review the model validation report.
＂ Apply the model to a dataflow entity.
＂ Use the scored output from the model in a Power BI report.

Create a dataflow with the input data

Create a dataflow with input data by following these steps.

Get data
The first step in creating a dataflow is to have your data sources ready. In this case, you
use a machine learning semantic model from a set of online sessions, some of which
culminated in a purchase. The semantic model contains a set of attributes about these
sessions, which you use to train your model.

You can download the semantic model from the UC Irvine website or by downloading
the online_shoppers_intention.csv . Later in this tutorial, you connect to the semantic
model by specifying its URL.

Create the tables

To create the entities in your dataflow, sign into the Power BI service and navigate to a
workspace.

1. If you don't have a workspace, create one by selecting Workspaces in the Power BI
left navigation pane and selecting Create a workspace. In the Create a workspace
panel, enter a workspace name and select Save.

2. Select New at the top of the new workspace, and then select Dataflow.
 

3. Select Add new tables to launch a Power Query editor in the browser.

4. On the Choose data source screen, select Text/CSV as the data source.
 

5. On the Connect to a data source page, paste the following link to the
online_shoppers_intention.csv file into the File path or URL box, and then select
Next.

https://raw.githubusercontent.com/santoshc1/PowerBI-AI-
samples/master/Tutorial_AutomatedML/online_shoppers_intention.csv
 

6. The Power Query Editor shows a preview of the data from the CSV file. To make
changes in the data before loading it, select Transform data.

7. Power Query automatically infers the data types of the columns. You can change
the data types by selecting the attribute type icon at the tops of the column
headers. Change the type of the Revenue column to True/False.
 You can rename the query to a friendlier name by changing the value in the Name
box in the right pane. Change the query name to Online visitors.

8. Select Save & close, and in the dialog box, provide a name for the dataflow and
then select Save.
 

Create and train a machine learning model

To add a machine learning model:

1. Select the Apply ML model icon in the Actions list for the table that contains your
training data and label information, and then select Add a machine learning
model.

2. The first step to create your machine learning model is to identify the historical
data, including the outcome field that you want to predict. The model is created by
learning from this data. In this case, you want to predict whether or not visitors are
 going to make a purchase. The outcome you want to predict is in the Revenue
field. Select Revenue as the Outcome column value, and then select Next.

3. Next, you select the type of machine learning model to create. Power BI analyzes
the values in the outcome field that you identified, and suggests the types of
machine learning models that it can create to predict that field.

In this case, since you want to predict a binary outcome of whether or not a visitor
is going to make a purchase, Power BI recommends Binary Prediction. Because
you're interested in predicting visitors who are going to make a purchase, select
true under Choose a target outcome. You can also provide different labels to use
for the outcomes in the automatically generated report that summarizes the model
validation results. Then select Next.
 

4. Power BI does a preliminary scan of a sample of your data and suggests inputs that
might produce more accurate predictions. If Power BI doesn't recommend a
column, it explains why not next to the column. You can change the selections to
include only the fields you want the model to study by selecting or deselecting the
checkboxes next to column names. Select Next to accept the inputs.
 

5. In the final step, name the model Purchase intent prediction, and choose the
amount of time to spend in training. You can reduce the training time to see quick
results or increase the time to get the best model. Then select Save and train to
start training the model.
 

If you get an error similar to Credentials not found for data source, you need to update
your credentials so Power BI can score the data. To update your credentials, select More
options ... in the header bar and then select Settings > Settings.

Select your dataflow under Dataflows, expand Data source credentials, and then select
Edit credentials.
 

Track training status

The training process begins by sampling and normalizing your historical data and
splitting your semantic model into two new entities: Purchase Intent Prediction Training
Data and Purchase Intent Prediction Testing Data.

Depending on the size of the semantic model, the training process can take anywhere
from a few minutes up to the training time you selected. You can confirm that the model
is being trained and validated through the status of the dataflow. The status appears as
a data refresh in progress in the Semantic models + dataflows tab of the workspace.

You can see the model in the Machine learning models tab of the dataflow. Status
indicates whether the model has been queued for training, is under training, or is
trained. Once the model training is completed, the dataflow displays an updated Last
trained time and a status of Trained.

Review the model validation report

To review the model validation report, in the Machine learning models tab, select the
View training report icon under Actions. This report describes how your machine
learning model is likely to perform.

In the Model Performance page of the report, select See top predictors to view the top
predictors for your model. You can select one of the predictors to see how the outcome
distribution is associated with that predictor.
 

You can use the Probability Threshold slicer on the Model Performance page to
examine the influence of model Precision and Recall on the model.

The other pages of the report describe the statistical performance metrics for the model.

The report also includes a Training Details page that describes the Iterations run, how
features were extracted from the inputs, and the hyperparameters for the Final model
used.

Apply the model to a dataflow entity

Select the Apply model button at the top of the report to invoke this model. In the
Apply dialog, you can specify the target entity that has the source data to apply the
model to. Then select Save and apply.

Applying the model creates two new tables, with the suffixes enriched <model_name>
and enriched <model_name> explanations. In this case, applying the model to the
Online visitors table creates:

Online visitors enriched Purchase intent prediction, which includes the predicted
output from the model.
Online visitors enriched Purchase intent prediction explanations, which contains
top record-specific influencers for the prediction.

Applying the binary prediction model adds four columns: Outcome, PredictionScore,
PredictionExplanation, and ExplanationIndex, each with a Purchase intent prediction
prefix.
 

Once the dataflow refresh completes, you can select the Online visitors enriched
Purchase intent prediction table to view the results.
 

You can also invoke any automated machine learning model in the workspace directly
from the Power Query Editor in your dataflow. To access the automated machine
learning models, select Edit for the table that you want to enrich with insights from your
automated machine learning model.

In the Power Query Editor, select AI insights in the ribbon.

 

On the AI insights screen, select the Power BI Machine Learning Models folder from
the navigation pane. The list shows all the machine learning models you have access to
as Power Query functions. The input parameters for the machine learning model
automatically map as parameters of the corresponding Power Query function. The
automatic parameter mapping happens only if the names and data types of the
parameter are the same.

To invoke a machine learning model, you can select any of the selected model's columns
as an input in the dropdown list. You can also specify a constant value to use as an input
by toggling the column icon next to the input line.

Select Apply to view the preview of the machine learning model output as new columns
in the table. You also see the model invocation under Applied steps for the query.
 

After you save your dataflow, the model automatically invokes when the dataflow
refreshes, for any new or updated rows in the entity table.

Using the scored output from the model in a

Power BI report
To use the scored output from your machine learning model, you can connect to your
dataflow from Power BI Desktop by using the Dataflows connector. You can now use the
Online visitors enriched Purchase intent prediction table to incorporate the predictions
from your model in Power BI reports.

Limitations
There are some known issues with using gateways with automated machine learning. If
you need to use a gateway, it's best to create a dataflow that imports the necessary data
via the gateway first. Then create another dataflow that references the first dataflow to
create or apply these models.

If your AI work with dataflows fails, you may need to enable Fast Combine when using
AI with dataflows. Once you have imported your table and before you begin to add AI
features, select Options from the Home ribbon, and in the window that appears select
the checkbox beside Allow combining data from multiple sources to enable the feature,
then select OK to save your selection. Then you can add AI features to your dataflow.

Related content
In this tutorial, you created and applied a binary prediction model in Power BI by doing
these steps:

Created a dataflow with the input data.

Created and trained a machine learning model.
Reviewed the model validation report.
 Applied the model to a dataflow entity.
Learned how to use the scored output from the model in a Power BI report.

For more information about Machine Learning automation in Power BI, see Automated
machine learning in Power BI.

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Refresh data from an on-premises SQL
Server database
Article • 05/28/2024

In this tutorial, you explore how to refresh a Power BI dataset from a relational database
that exists on premises in your local network. Specifically, this tutorial uses a sample SQL
Server database, which Power BI must access through an on-premises data gateway.

In this tutorial, you complete the following steps:

＂ Create and publish a Power BI Desktop .pbix file that imports data from an on-
premises SQL Server database.
＂ Configure data source and dataset settings in Power BI for SQL Server connectivity
through a data gateway.
＂ Configure a refresh schedule to ensure your Power BI dataset has recent data.
＂ Do an on-demand refresh of your dataset.
＂ Review the refresh history to analyze the outcomes of past refresh cycles.
＂ Clean up resources by deleting the items you created in this tutorial.

Prerequisites
If you don't already have one, sign up for a free Power BI trial before you begin.
Install Power BI Desktop on a local computer.
Install SQL Server on a local computer, and restore the AdventureWorksDW2017
sample database from a backup . For more information about the
AdventureWorks sample databases, see AdventureWorks installation and
configuration.
Install SQL Server Management Studio (SSMS).
Install an on-premises data gateway on the same local computer as SQL Server. In
production, the gateway would usually be on a different computer.

７ Note

If you're not a gateway administrator, or don't want to install a gateway yourself,

ask a gateway administrator in your organization to create the required data source
definition to connect your dataset to your SQL Server database.

Create and publish a Power BI Desktop file

Use the following procedure to create a basic Power BI report that uses the
AdventureWorksDW2017 sample database. Publish the report to the Power BI service to
get a Power BI dataset, which you configure and refreshes in later steps.

1. In Power BI Desktop, on the Home tab, select Get data > SQL Server.

2. In the SQL Server database dialog box, enter the Server and Database (optional)
names, and make sure the Data Connectivity mode is set to Import.

７ Note

If you plan to use a stored procedure, you must use Import as the Data
connectivity mode.

Optionally, under Advanced options, you could specify a SQL statement and set
other options like using SQL Server Failover.
3. Select OK.

4. On the next screen, verify your credentials, and then select Connect.

７ Note

If authentication fails, make sure you selected the correct authentication

method and used an account with database access. In test environments, you
might use Database authentication with an explicit username and password.
In production environments, you typically use Windows authentication. For
more assistance, see Troubleshoot refresh scenarios, or contact your
database administrator.

5. If an Encryption Support dialog box appears, select OK.

6. In the Navigator dialog box, select the DimProduct table, and then select Load.
7. In the Power BI Desktop Report view, in the Visualizations pane, select the Stacked
column chart.

8. With the new column chart selected in the report canvas, in the Fields pane, select
the EnglishProductName and ListPrice fields.
9. Drag EndDate from the Fields pane onto Filters on this page in the Filters pane,
and under Basic filtering, select the checkbox for (Blank).

The visualization should now look similar to the following chart:

 Notice that the Road-250 Red product has the same list price as the other Road-
250 products. This price changes when you later update the data and refresh the
report.

10. Save the report with the name AdventureWorksProducts.pbix.

11. On the Home tab, select Publish.

12. On the Publish to Power BI screen, choose My Workspace, and then select Select.
Sign in to the Power BI service if necessary.

13. When the Success message appears, select Open 'AdventureWorksProducts.pbix'

in Power BI.
Connect the dataset to the SQL Server
database
In Power BI Desktop, you connected directly to your on-premises SQL Server database.
In the Power BI service, you need a data gateway to act as a bridge between the cloud
and your on-premises network. Follow these steps to add your on-premises SQL Server
database as a data source to a gateway and connect your dataset to this data source.

1. In the Power BI service, in the upper-right corner of the screen, select the settings
gear icon and then select Settings.
2. Select the Semantic models tab, and then select the AdventureWorksProducts
dataset from the list of datasets.

3. Expand Gateway connection and verify that at least one gateway is listed. If you
don't see a gateway, make sure you followed the instructions to install an on-
premises data gateway.

4. Select the arrow toggle under Actions to expand the data sources, and then select
the Add to gateway link next to your data source.

5. On the New connection screen with On-premises selected, complete or verify the
following fields. Most fields are already filled in.

Gateway cluster name: Verify or enter the gateway cluster name.

Connection name: Enter a name for the new connection, such as
AdventureWorksProducts.
Connection type: Select SQL Server if not already selected.
 Server: Verify or enter your SQL Server instance name. Must be identical to
what you specified in Power BI Desktop.
Database: Verify or enter your SQL Server database name, such as
AdventureWorksDW2017. Must be identical to what you specified in Power
BI Desktop.

Under Authentication:

Authentication method: Select Windows, Basic, or OAuth2, usually

Windows.
Username and Password: Enter the credentials you use to connect to SQL
Server.
6. Select Create.

7. Back on the Settings screen, expand the Gateway connection section, and verify
that the data gateway you configured now shows a Status of running on the
machine where you installed it. Select Apply.
Configure a refresh schedule
Once connected, your Power BI dataset to your SQL Server on-premises database
through a data gateway, follow these steps to configure a refresh schedule. Refreshing
your dataset on a scheduled basis helps ensure that your reports and dashboards have
the most recent data.

1. In the left navigation pane, expand My Workspace.

2. In the Semantic models section, point to the AdventureWorksProducts dataset,

select the Open menu three vertical dots icon, and then select Schedule refresh.

 Tip

Make sure you point to the AdventureWorksProducts dataset, not the report
with the same name, which doesn't have a Schedule refresh option.

3. In the Scheduled refresh section, under Keep your data up to date, set refresh to
On.

4. Under Refresh frequency, select Daily for this example, and then under Time,
select Add another time.

For this example, specify 6:00 AM, then select Add another time and specify 6:00
PM.
 ７ Note

You can configure up to eight daily time slots if your dataset is on shared
capacity, or 48 time slots on Power BI Premium.

5. Leave the checkbox under Send refresh failure notifications to set to Semantic
model owner, and select Apply.

With a configured refresh schedule, Power BI refreshes your dataset at the next
scheduled time, within a margin of 15 minutes.

Do an on-demand refresh
To refresh the data anytime, such as to test your gateway and data source configuration,
you can do an on-demand refresh by using the Refresh Now option in the left pane
Semantic model menu. On-demand refreshes don't affect the next scheduled refresh
time.

To illustrate an on-demand refresh, first change the sample data by using SSMS to
update the DimProduct table in the AdventureWorksDW2017 database, as follows:

SQL

UPDATE [AdventureWorksDW2017].[dbo].[DimProduct]
SET ListPrice = 5000
WHERE EnglishProductName ='Road-250 Red, 58'

Follow these steps to make the updated data flow through the gateway connection to
the dataset and into the Power BI reports:

1. In the Power BI service, expand My Workspace in the left navigation pane.

2. In the Semantic models section, hover over the AdventureWorksProducts dataset,

select the three vertical dots Open menu icon, and then select Refresh now.

A Preparing for refresh message appears at upper right.

3. In the Reports section of My Workspace, select AdventureWorksProducts. See

how the updated data flowed through into the report, and the product with the
highest list price is now Road-250 Red, 58.
Review the refresh history
It's a good idea to periodically use the refresh history to check the outcomes of past
refresh cycles. Database credentials might have expired, or the selected gateway might
have been offline when a scheduled refresh was due. Follow these steps to examine the
refresh history and check for issues.

1. In the upper-right corner of the Power BI screen, select the settings gear icon and
then select Settings.

2. On the Semantic models tab, select the dataset you want to examine, such as
AdventureWorksProducts.

3. Select the Refresh history link.

4. On the Scheduled tab of the Refresh history dialog box, notice the past scheduled
and on-demand refreshes with their Start and End times. A Status of Completed
indicates that Power BI did the refreshes successfully. For failed refreshes, you can
see the error message and examine error details.
 ７ Note

The OneDrive tab is relevant only for datasets that are connected to Power BI
Desktop files, Excel workbooks, or CSV files on OneDrive or SharePoint Online.
For more information, see Data refresh in Power BI.

Clean up resources
Follow these instructions to clean up the resources you created for this tutorial:

If you don't want to use the sample data anymore, use SSMS to drop the database.
If you don't want to use the SQL Server data source, remove the data source from
your data gateway. Also consider uninstalling the data gateway, if you installed it
only for this tutorial.
Also delete the AdventureWorksProducts dataset and report that Power BI created
when you published the AdventureWorksProducts.pbix file.

Related content
This tutorial explored how to:

Import data from an on-premises SQL Server database into a Power BI dataset.
To update reports and dashboards that use the dataset, refresh the Power BI
dataset on a scheduled and on-demand basis.

Now, you can learn more about Power BI data refresh and managing data gateways and
data sources.

Manage an on-premises data gateway

Manage your data source - Import/scheduled refresh
 Data refresh in Power BI

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Tutorial: Automate configuration of
template app installation using an Azure
function
Article • 12/21/2023

Template apps are a great way for customers to start getting insights from their data.
Template apps get them up and running quickly by connecting them to their data. The
template apps provide customers with prebuilt reports that they can customize if they
so desire.

Customers aren't always familiar with the details of how to connect to their data. Having
to provide these details when they install a template app can be a pain point for them.

If you are a data services provider and have created a template app to help your
customers get started with their data on your service, you can make it easier for them to
install your template app. You can automate the configuration of your template app's
parameters.

When the customer signs in to your portal, they select a special link you've prepared.
This link:

Launches the automation, which gathers the information it needs.

Preconfigures the template app parameters.
Redirects the customer to their Power BI account where they can install the app.

All they have to do is select Install and authenticate against their data source, and
they're good to go!

The customer experience is illustrated here.

In this tutorial, you'll use an automated installation Azure Functions sample that we've
created to preconfigure and install your template app. This sample has deliberately been
kept simple for demonstration purposes. It encapsulates the setup of an Azure function
to use Power BI APIs for installing a template app and configuring it for your users
automatically.

For more information about the general automation flow and the APIs that the app
uses, see Automate configuration of a template app installation.

Our simple application uses an Azure function. For more information about Azure
Functions, see the Azure Functions documentation.

Basic flow
The following basic flow lists what the application does when the customer launches it
by selecting the link in your portal.

1. The user signs in to the ISV's portal and selects the supplied link. This action
initiates the flow. The ISV's portal prepares the user-specific configuration at this
stage.

2. The ISV acquires an app-only token based on a service principal (app-only token)
that's registered in the ISV's tenant.

3. Using Power BI REST APIs, the ISV creates an install ticket, which contains the user-
specific parameter configuration as prepared by the ISV.

4. The ISV redirects the user to Power BI by using a POST redirection method, which
contains the install ticket.
 5. The user is redirected to their Power BI account with the install ticket and is
prompted to install the template app. When the user selects Install, the template
app is installed for them.

７ Note

While parameter values are configured by the ISV in the process of creating the
install ticket, data source-related credentials are only supplied by the user in the
final stages of the installation. This arrangement prevents them from being exposed
to a third party and ensures a secure connection between the user and the
template app data sources.

Prerequisites
Your own Microsoft Entra tenant set up. For instructions on how to set one up, see
Create a Microsoft Entra tenant.
A service principal (app-only token) registered in the preceding tenant.
A parameterized template app that's ready for installation. The template app must
be created in the same tenant in which you register your application in Microsoft
Entra ID. For more information, see Template app tips or Create a template app in
Power BI.
To be able to test your automation work flow, add the service principal to the
template app workspace as an Admin.
A Power BI Pro license. If you're not signed up for Power BI Pro, sign up for a free
trial before you begin.

Set up your template apps automation

development environment
Before you continue setting up your application, follow the instructions in Quickstart:
Create an Azure Functions app with Azure App Configuration to develop an Azure
function along with an Azure app configuration. Create your app configuration as
described in the article.

Register an application in Microsoft Entra ID

Create a service principal as described in Embed Power BI content with service principal
and an application secret.
Make sure to register the application as a server-side web application app. You register
a server-side web application to create an application secret.

Save the application ID (ClientID) and application secret (ClientSecret) for later steps.

You can go through the Embedding setup tool to quickly get started creating an app
registration. If you're using the Power BI App Registration Tool , select the Embed for
your customers option.

Add the service principal to the template app workspace as an Admin, so that you will
be able to test your automation work flow.

Template app preparation

After you've created your template app and it's ready for installation, save the following
information for the next steps:

App ID, Package Key, and Owner ID as they appear in the installation URL at the
end of the Define the properties of the template app process when the app was
created.

You can also get the same link by selecting Get link in the template app's Release
Management pane.

Parameter names as they're defined in the template app's semantic model.

Parameter names are case-sensitive strings. They can also be retrieved from the
Parameter Settings tab when you define the properties of the template app or
from the semantic model settings in Power BI.

７ Note

You can test your preconfigured installation application on your template app if the
template app is ready for installation, even if it isn't publicly available on AppSource
yet. For users outside your tenant to be able to use the automated installation
application to install your template app, the template app must be publicly
available in the Power BI apps marketplace . Before you distribute your template
app by using the automated installation application you're creating, be sure to
publish it to Partner Center.

Install and configure your template app

In this section, you'll use an automated installation Azure Functions sample that we
created to preconfigure and install your template app. This sample has deliberately been
kept simple for demonstration purposes. It allows you to use an Azure function and
Azure App Configuration to easily deploy and use the automated installation API for
your template apps.

Download Visual Studio (version 2017 or later)

Download Visual Studio (version 2017 or later). Make sure to download the latest
NuGet package .

Download the automated installation Azure Functions

sample
Download the automated installation Azure Functions sample from GitHub to get
started.

Set up your Azure app configuration

To run this sample, you need to set up your Azure app configuration with the values and
keys as described here. The keys are the application ID, the application secret, and your
template app's AppId, PackageKey, and OwnerId values. See the following sections for
information about how to get these values.

The keys are also defined in the Constants.cs file.

ﾉ Expand table
 Configuration key Meaning

TemplateAppInstall:Application:AppId AppId from the installation URL

TemplateAppInstall:Application:PackageKey PackageKey from the installation URL

TemplateAppInstall:Application:OwnerId OwnerId from the installation URL

TemplateAppInstall:ServicePrincipal:ClientId Service principal application ID

TemplateAppInstall:ServicePrincipal:ClientSecret Service principal application secret

The Constants.cs file is shown here.

Get the template app properties

Fill in all relevant template app properties as they're defined when the app is created.
These properties are the template app's AppId, PackageKey, and OwnerId values.

To get the preceding values, follow these steps:

1. Sign in to Power BI .

2. Go to the application's original workspace.

3. Open the Release Management pane.

4. Select the app version, and get its installation link.

5. Copy the link to the clipboard.

 6. This installation URL holds the three URL parameters whose values you need. Use
the appId, packageKey, and ownerId values for the application. A sample URL will
be similar to what is shown here.

HTML

https://app.powerbi.com/Redirect?
action=InstallApp&appId=3c386...16bf71c67&packageKey=b2df4b...dLpHIUnum
2pr6k&ownerId=72f9...1db47&buildVersion=5

Get the application ID

Fill in the applicationId information with the application ID from Azure. The
applicationId value is used by the application to identify itself to the users from which
you're requesting permissions.

To get the application ID, follow these steps:

1. Sign in to the Azure portal .

2. In the left pane, select All services > App registrations.

3. Select the application that needs the application ID.

4. There's an application ID that's listed as a GUID. Use this application ID as the

applicationId value for the application.
Get the application secret
Fill in the ApplicationSecret information from the Keys section of your App
registrations section in Azure. This attribute works when you use the service principal.

To get the application secret, follow these steps:

1. Sign in to the Azure portal .

2. In the left pane, select All services > App registrations.

3. Select the application that needs to use the application secret.

4. Select Certificates and secrets under Manage.

5. Select New client secrets.

6. Enter a name in the Description box, and select a duration. Then select Save to get
the value for your application. When you close the Keys pane after you save the
key value, the Value field shows only as hidden. At that point, you aren't able to
retrieve the key value. If you lose the key value, create a new one in the Azure
portal.
Test your function locally
Follow the steps as described in Run the function locally to run your function.

Configure your portal to issue a POST request to the URL of the function. An example is
POST http://localhost:7071/api/install . The request body should be a JSON object

that describes key-value pairs. Keys are parameter names as defined in Power BI
Desktop. Values are the desired values to be set for each parameter in the template app.

７ Note

In production, parameter values are deduced for each user by your portal's
intended logic.

The desired flow should be:

1. The portal prepares the request, per user or session.

2. The POST /api/install request is issued to your Azure function. The request body
consists of key-value pairs. The key is the parameter name. The value is the desired
value to be set.
3. If everything is configured properly, the browser should automatically redirect to
the customer's Power BI account and show the automated installation flow.
4. Upon installation, parameter values are set as configured in steps 1 and 2.

Next steps

Publish your project to Azure

To publish your project to Azure, follow the instructions in the Azure Functions
documentation. Then you can integrate template app automated installation APIs into
your product and begin testing it in production environments.
Power BI data sources
Article • 06/09/2023

Power BI uses Power Query to connect to data sources. Power BI data sources are
documented in the following article: Power Query (including Power BI) connectors.

Each data source article in the Power Query documentation describes the capabilities of
the data connector, such as whether DirectQuery is supported. The following image
shows the Capabilities supported section for Azure Data Explorer (Kusto), where it
states that DirectQuery is supported for the connector in Power BI. If DirectQuery (or any
other capability) isn't listed, the capability isn't supported.

For a list of the connectors available in Power Query, see connectors in Power Query.

For information about dataflows in Power BI, see connect to data sources for Power BI
dataflows.

Considerations and limitations

Many data connectors for Power BI Desktop require Internet Explorer 10 (or newer)
for authentication.
Some data sources are available in Power BI Desktop optimized for Power BI
Report Server, but aren't supported when published to Power BI Report Server. See
Power BI report data sources in Power BI Report Server for the list of supported
data sources.
Power BI Desktop and the Power BI service may send multiple queries for any
given query, to get schema information or the data itself, based in part on whether
 data is cached. This behavior is by design, for more information, see the Power
Query article that describes why a query may run multiple times.

Next steps
The following articles provide more information about Power BI and connecting to data:

Connectors in Power Query

Connect to data in Power BI Desktop
Using DirectQuery in Power BI
What is an on-premises data gateway?
Power BI report data sources in Power BI Report Server
New name for Power BI datasets
Article • 11/14/2023

Microsoft has renamed the Power BI dataset content type to semantic model.

The rename was necessary for two main reasons.

The term dataset is considered too generic. It has different meanings in the context
of other data-related activities, especially now that Power BI is one of many
experiences in Microsoft Fabric.
The term semantic model better reflects the rich functionality of Analysis Services
data models, upon which Power BI reports are based.

） Important

This change is a rename only. There's no interruption to usage or service. You can
expect a continuation of service because administrators, developers, and other
users aren't required to make any changes.

 Tip

To avoid confusion and support requests, be sure to notify your community of

practice of this change.

Name changes
Here are some examples of name changes.

ﾉ Expand table

Old name New name

Dataset Semantic model

Shared dataset Shared semantic model

Import dataset Import semantic model

DirectQuery dataset DirectQuery semantic model

Composite dataset Composite semantic model

 Old name New name

Live connection dataset Live connection semantic model

On-premises dataset On-premises semantic model

Dataset owner Semantic model owner

Large dataset Large semantic model

７ Note

The name change has been rolled out in the Power BI service and in
documentation, though there might be some instances where the change hasn't
occurred yet.

Name change exceptions

The following concepts aren't affected.

Generic references to datasets

Power BI paginated report dataset
Power BI real-time dataset, including:
Push dataset
Streaming dataset
Hybrid dataset
PubNub dataset
All Power BI REST API operations related to datasets
Power BI activity log operations
Other types of datasets that aren't related to Power BI, for example, Azure Open
Datasets

Related content
For more information related to this article, check out the following resources.

Blog post: Datasets renamed to semantic models

Questions? Try asking the Fabric Community .
Suggestions? Contribute ideas to improve Fabric .
Semantic models in the Power BI service
Article • 11/10/2023

This article provides a technical explanation of Power BI semantic models.

Semantic model types

Power BI semantic models represent a source of data that's ready for reporting and
visualization. You can create Power BI semantic models in the following ways:

Connect to an existing data model that isn't hosted in Power BI.

Upload a Power BI Desktop file that contains a model.
Upload an Excel workbook that contains one or more Excel tables and/or a
workbook data model, or upload a comma-separated values (CSV) file.
Use the Power BI service to create a push semantic model.
Use the Power BI service to create a streaming or hybrid streaming semantic
model.

Except for streaming semantic models, semantic models represent data models, which
use the mature modeling technologies of Analysis Services.

７ Note

Power BI documentation sometimes uses the terms semantic model and model
interchangeably. A semantic model in the Power BI service refers to a model from a
development perspective. In a documentation context, the terms mean much the
same thing.

External-hosted models
There are two types of external-hosted models: SQL Server Analysis Services and Azure
Analysis Services.

To connect to a SQL Server Analysis Services model, you must install an on-premises
data gateway either on-premises or on a virtual machine-hosted infrastructure-as-a-
service (IaaS). Azure Analysis Services doesn't require a gateway.

It often makes sense to connect to Analysis Services when there are existing model
investments, which typically form part of an enterprise data warehouse (EDW). Power BI
can make a live connection to Analysis Services, and enforce data permissions by using
the identity of the Power BI report user.

SQL Server Analysis Services supports both multidimensional models, or cubes, and
tabular models. As the following image shows, a live connection semantic model passes
queries to external-hosted models.

Power BI Desktop-developed models

You can use Power BI Desktop, a client application for Power BI development, to develop
a model. A Power BI Desktop model is effectively an Analysis Services tabular model.

You can develop three different types, or modes, of models by using Power BI Desktop:
Import, DirectQuery, and Composite. You develop models by importing data from
dataflows and then integrating them with external data sources. The mode depends on
whether data is imported into the model, or whether it remains in the data source. For
more information about the modes, see Semantic model modes in the Power BI service.

Semantic model ownership

When working with semantic models using gateway and cloud connections, your ability
to make changes to the semantic model is dependent on ownership of the semantic
model. If you're not the owner, a warning is displayed stating that you're viewing the
section of the semantic model information in read-only mode because you're not the
semantic model owner. To make changes, you must either contact the semantic model
owner to make changes, or take over ownership of the semantic model.

Row-level security
External-hosted models and Power BI desktop models can enforce row-level security
(RLS) to limit the data that certain users can retrieve. For example, users assigned to a
Salespeople security group might be able to view report data only for the sales regions
they're assigned to. RLS roles are dynamic or static. Dynamic roles filter by the report
user, while static roles apply the same filters for all users assigned to the role. For more
information, see Row-level security (RLS) with Power BI.

Excel workbook models

Creating semantic models based on Excel workbooks or CSV files automatically creates a
model. Imported Excel tables and CSV data create model tables, while Excel workbook
data transposes to create a Power BI model. In all cases, file data imports into a model.

Summary
In summary:

Power BI semantic models that represent models are either hosted in the Power BI
service, or are externally hosted by Analysis Services.
Semantic model models can store imported data, or issue pass-through query
requests to underlying data sources, or do both.

Considerations
The following important facts and considerations apply to Power BI semantic models
that represent models:

SQL Server Analysis Services-hosted models need a gateway to do live connection

queries.
To query Power BI-hosted models that import data, you must fully load them into
memory.
Power BI-hosted models that use Import need refresh to keep data current, and
must use gateways when source data isn't accessible directly over the internet.
Power BI-hosted Import models can refresh according to a schedule, or a user can
trigger on-demand refresh in the Power BI service.
Power BI-hosted models that use DirectQuery mode require connectivity to the
source data. Power BI issues queries to the source data to retrieve current data.
This mode must use gateways when source data isn't accessible directly over the
internet.
Models can enforce RLS rules to filter data access to certain users.
You can use the semantic models - Take Over In Group API to take over ownership
if a semantic model owner leaves the organization.
To successfully deploy and manage Power BI semantic models, you should understand
the following factors:

The model design itself, including its data preparation queries, relationships, and
calculations.
The following configurations that can significantly impact Power BI capacity
resources:
Where models are hosted
The storage mode
Any dependencies on gateways
The size of imported data
Model refresh type and frequency

Next steps
Semantic model modes in the Power BI service
Questions? Ask the Power BI Community
Suggestions? Contribute ideas to improve Power BI
Semantic model modes in the Power BI
service
Article • 11/10/2023

This article provides a technical explanation of Power BI semantic model modes. It

applies to semantic models that represent a live connection to an external-hosted
Analysis Services model, and also to models developed in Power BI Desktop. The article
emphasizes the rationale for each mode, and possible impacts on Power BI capacity
resources.

The three semantic model modes are:

Import
DirectQuery
Composite

Import mode
Import mode is the most common mode used to develop semantic models. This mode
delivers fast performance thanks to in-memory querying. It also offers design flexibility
to modelers, and support for specific Power BI service features (Q&A, Quick Insights,
etc.). Because of these strengths, it's the default mode when creating a new Power BI
Desktop solution.

It's important to understand that imported data is always stored to disk. When queried
or refreshed, the data must be fully loaded into memory of the Power BI capacity. Once
in memory, Import models can then achieve very fast query results. It's also important to
understand that there's no concept of an Import model being partially loaded into
memory.

When refreshed, data is compressed and optimized and then stored to disk by the
VertiPaq storage engine. When loaded from disk into memory, it's possible to see 10-
times compression. So, it's reasonable to expect that 10 GB of source data can compress
to about 1 GB in size. Storage size on disk can achieve a 20% reduction from the
compressed size. The difference in size can be determined by comparing the Power BI
Desktop file size with the Task Manager memory usage of the file.

Design flexibility can be achieved in three ways:

Integrate data by caching data from dataflows, and external data sources, whatever
the data source type or format.
 Use the entire set of Power Query M formula language, referred to as M, functions
when creating data preparation queries.
Apply the entire set of Data Analysis Expressions (DAX) functions when enhancing
the model with business logic. There's support for calculated columns, calculated
tables, and measures.

As shown in the following image, an Import model can integrate data from any number
of supported data source types.

However, while there are compelling advantages associated with Import models, there
are disadvantages, too:

The entire model must be loaded to memory before Power BI can query the
model, which can place pressure on available capacity resources, especially as the
number and size of Import models grow.
Model data is only as current as the latest refresh, and so Import models need to
be refreshed, usually on a scheduled basis.
A full refresh removes all data from all tables and reloads it from the data source.
This operation can be expensive in terms of time and resources for the Power BI
service, and the data sources.

７ Note

Power BI can achieve incremental refresh to avoid truncating and reloading entire
tables. For more information, including supported plans and licensing, see
Incremental refresh and real-time data for semantic models.
From a Power BI service resource perspective, Import models require:

Sufficient memory to load the model when it's queried or refreshed.

Processing resources and extra memory resources to refresh data.

DirectQuery mode
DirectQuery mode is an alternative to Import mode. Models developed in DirectQuery
mode don't import data. Instead, they consist only of metadata defining the model
structure. When the model is queried, native queries are used to retrieve data from the
underlying data source.

There are two main reasons to consider developing a DirectQuery model:

When data volumes are too large, even when data reduction methods are applied,
to load into a model, or practically refresh.
When reports and dashboards need to deliver near real-time data, beyond what
can be achieved within scheduled refresh limits. Scheduled refresh limits are eight
times a day for shared capacity, and 48 times a day for a Premium capacity.

There are several advantages associated with DirectQuery models:

Import model size limits don't apply.

Models don't require scheduled data refresh.
Report users see the latest data when interacting with report filters and slicers.
Also, report users can refresh the entire report to retrieve current data.
Real-time reports can be developed by using the Automatic page refresh feature.
Dashboard tiles, when based on DirectQuery models, can update automatically as
frequently as every 15 minutes.

However, there are some limitations associated with DirectQuery models:

Power Query/Mashup expressions can only be functions that can be transposed to

native queries understood by the data source.
DAX formulas are limited to use only functions that can be transposed to native
queries understood by the data source. Calculated tables aren't supported.
 Quick Insights features aren't supported.

From a Power BI service resource perspective, DirectQuery models require:

Minimal memory to load the model (metadata only) when it's queried.
Sometimes the Power BI service must use significant processor resources to
generate and process queries sent to the data source. When this situation arises, it
can affect throughput, especially when concurrent users are querying the model.

For more information, see Use DirectQuery in Power BI Desktop.

Composite mode
Composite mode can mix Import and DirectQuery modes, or integrate multiple
DirectQuery data sources. Models developed in Composite mode support configuring
the storage mode for each model table. This mode also supports calculated tables,
defined with DAX.

The table storage mode can be configured as Import, DirectQuery, or Dual. A table
configured as Dual storage mode is both Import and DirectQuery, and this setting
allows the Power BI service to determine the most efficient mode to use on a query-by-
query basis.

Composite models strive to deliver the best of Import and DirectQuery modes. When
configured appropriately, they can combine the high query performance of in-memory
models with the ability to retrieve near real-time data from data sources.
For more information, see Use composite models in Power BI Desktop.

Pure Import and DirectQuery tables

Data modelers who develop Composite models are likely to configure dimension-type
tables in Import or Dual storage mode, and fact-type tables in DirectQuery mode. For
more information about model table roles, see Understand star schema and the
importance for Power BI.

For example, consider a model with a Product dimension-type table in Dual mode, and
a Sales fact-type table in DirectQuery mode. The Product table could be efficiently and
quickly queried from in-memory to render a report slicer. The Sales table could also be
queried in DirectQuery mode with the related Product table. The latter query could
enable the generation of a single efficient native SQL query that joins Product and Sales
tables, and filters by the slicer values.

Hybrid tables
Data modelers who develop Composite models can also configure fact tables as hybrid
tables. A hybrid table is a table with one or multiple Import partitions and one
DirectQuery partition. The advantage of a hybrid table is that it could be efficiently and
quickly queried from in-memory while at the same time including the latest data
changes from the data source that occurred after the last import cycle, as the following
visualization illustrates.
The easiest way to create a hybrid table is to configure an incremental refresh policy in
Power BI Desktop and enable the option Get the latest data in real time with
DirectQuery (Premium only). When Power BI applies an incremental refresh policy that
has this option enabled, it partitions the table like the partitioning scheme displayed in
the previous diagram. To ensure good performance, configure your dimension-type
tables in Dual storage mode so that Power BI can generate efficient native SQL queries
when querying the DirectQuery partition.

７ Note

Power BI supports hybrid tables only when the semantic model is hosted in
workspaces on Premium capacities. Accordingly, you must upload your semantic
model to a Premium workspace if you configure an incremental refresh policy with
the option to get the latest data in real time with DirectQuery. For more
information, see Incremental refresh and real-time data for semantic models.

It's also possible to convert an Import table to a hybrid table by adding a DirectQuery
partition using Tabular Model Scripting Language (TMSL) or the Tabular Object Model
(TOM) or by using a third-party tool. For example, you can partition a fact table such
that the bulk of the data is left in the data warehouse while only a fraction of the most
recent data is imported. This approach can help to optimize performance if the bulk of
this data is historical data that is infrequently accessed. A hybrid table can have multiple
Import partitions, but only one DirectQuery partition.

Next steps
Storage mode in Power BI Desktop
Using DirectQuery in Power BI
Use composite models in Power BI Desktop
More questions? Try asking the Power BI Community
Power BI data source prerequisites
Article • 02/08/2023

For data sources, Power BI supports specific provider versions and data source versions,
and certain objects. For more information about available Power BI data sources, see
Data sources.

The following table describes Power BI data source requirements.

Data Provider Minimum Minimum Supported Download

source provider data data source link
version source objects
version

SQL Server ADO.net (built into .NET .NET SQL Tables/Views, Included in
Framework) Framework Server Scalar .NET
3.5 (only) 2005+ functions, Framework
Table 3.5 or
functions above

Access Microsoft Access Database ACE 2010 No Tables/Views Download

Engine (ACE) SP1 restriction link

Excel (.xls Microsoft Access Database ACE 2010 No Tables, Download

files only) Engine (ACE) SP1 restriction Sheets link
(see note
1)

Oracle (see ODP.NET ODAC 11.2 9.x+ Tables/Views Download

note 2) Release 5 link
(11.2.0.3.20)

MySQL Connector/Net 6.6.5 5.1 Tables/Views, Download

Scalar link
functions

PostgreSQL NPGSQL ADO.NET provider 4.0.10 9.4 Tables/Views Download

(Shipped with Power BI link
Desktop)

Teradata .NET Data Provider for 14+ 12+ Tables/Views Download

Teradata link

SAP Sybase iAnywhere.Data.SQLAnywhere 16+ 16+ Tables/Views Download

SQL for .NET 3.5 link
Anywhere
７ Note

Excel files that have an .xlsx extension do not require a separate provider
installation.

７ Note

The Oracle providers also require Oracle client software (version 8.1.7+).
Using enhanced semantic model
metadata
Article • 11/10/2023

When Power BI Desktop creates reports, it also creates semantic model metadata in the
corresponding PBIX and PBIT files. Previously, the metadata was stored in a format that
was specific to Power BI Desktop. The metadata used base-64 encoded M expressions
and data sources. Power BI made assumptions about how that metadata was stored.

With the release of the enhanced semantic model metadata feature, many of these
limitations are removed. PBIX files are automatically upgraded to enhanced metadata
upon opening the file. With enhanced semantic model metadata, metadata created by
Power BI Desktop uses a format similar to what is used for Analysis Services tabular
models, based on the Tabular Object Model.

The enhanced semantic model metadata feature is strategic and foundational. Future
Power BI functionality will be built upon its metadata. These other capabilities stand to
benefit from enhanced semantic model metadata:

XMLA read/write for management of Power BI semantic models.

Migration of Analysis Services workloads to Power BI to benefit from next-
generation features.

Upgrade
Your reports are automatically upgraded to the enhanced metadata format when you
open them in the latest version of Power BI Desktop. If the report was saved with
unapplied query changes, or there was an error during the auto-upgrade, there's a
warning on the report canvas that you still need to upgrade. Selecting Upgrade report
applies any pending changes and upgrades the data model to the new format.

Exclude table from report refresh

Once a data model has been upgraded to the enhanced metadata format, some
metadata that was previously only used in Power BI Desktop is now respected in the
Power BI service as well. This metadata includes the Include in Report Refresh option.
For upgraded models, if the Include in Report Refresh option is unselected in the Power
Query Editor, then that table isn't refreshed when the report or semantic model is
refreshed in Power BI Desktop or the Power BI service. Reports already published in the
Power BI service that aren't yet upgraded to the new enhanced metadata formal need to
be upgraded in Power BI Desktop before this new behavior takes effect.

Considerations and limitations

Before enhanced metadata support, for SQL Server, Oracle, Teradata, and legacy HANA
connections, Power BI Desktop added a native query to the data model. This query is
used by the Power BI service data models. With enhanced metadata support, the Power
BI service data model regenerates the native query at runtime. It doesn't use the query
that Power BI Desktop created. In most cases, this retrieval resolves itself correctly, but
some transformations don't work without reading underlying data. You might see some
errors in reports that previously worked. For example, an error might say:

Unable to convert an M query in table 'Dimension City' into a native source query.
Try again later or contact support. If you contact support, provide these details.

You can fix your queries in three different places in Power BI Desktop:

When you apply changes or do a refresh.

In a warning bar in the Power Query Editor informing you that the expression
couldn’t be folded to the data source.

When you run evaluations when you open a report to check if you have
unsupported queries. Running these evaluations can result in performance
implications.

Certain character combinations in M expressions that would be unsupported in the

Tabular Object Model (TOM) are also unsupported in the enhanced semantic model
metadata environment.

Next steps
You can do all sorts of things with Power BI Desktop. For more information on its
capabilities, check out the following resources:

What is Power BI Desktop?

What's new in Power BI?
Query overview with Power BI Desktop
Data types in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop
Work with multidimensional models in
Power BI
Article • 01/19/2023

You can connect to multidimensional models in Power BI, and create reports that
visualize all sorts of data within the model. With multidimensional models, Power BI
applies rules to how it processes data, based on which column is defined as the default
member.

With multidimensional models, Power BI handles data from the model based on where
the column that contains the Default Member is used. The DefaultMember property
value for an attribute hierarchy is set in CSDL (Conceptual Schema Definition Language)
for a particular column in a multidimensional model. For more information about the
default member, see Attribute properties - Define a default member. When a data
analysis expression (DAX) query is executed, the default member specified in the model
is applied automatically.

This article describes how Power BI behaves under various circumstances when working
with multidimensional models, based on where the default member is found.

Work with filter cards

When you create a filter card on a field with a default member, the default member field
value is selected automatically in the filter card. The result is that all visuals affected by
the filter card retain their default models in the database. The values in such filter cards
reflect that default member.

If the default member is removed, deselecting the value clears it for all visuals to which
the filter card applies, and the values displayed don't reflect the default member.

For example, imagine we have a Currency column and a default member set to USD:

In this example case, if we have a card that shows Total Sales, the value will have
the default member applied and the sales that correspond to USD.
If we drag Currency to the filter card pane, we see USD as the default value
selected. The value of Total Sales remains the same, since the default member is
applied.
However, if we deselect the USD value from the filter card, the default member for
Currency is cleared, and now Total Sales reflects all currencies.
 When we select another value in the filter card (let's say we select EURO), along the
default member, the Total Sales reflects the filter Currency IN {USD, EURO}.

Group visuals
In Power BI, whenever you group a visual on a column that has a default member, Power
BI clears the default member for that column and its attribute relationship path. This
behavior ensures the visual displays all values, instead of just the default values.

Attribute relationship paths (ARPs)

Attribute relationship paths (ARPs) provide default members with powerful capabilities,
but also introduce a certain amount of complexity. When ARPs are encountered, Power
BI follows the path of ARPs to clear other default members for other columns to provide
consistent, and precise handling of data for visuals.

Let's look at an example to clarify the behavior. Consider the following configuration of
ARPs:

Now let's imagine that the following default members are set for these columns:

City > Seattle

State > WA
Country/Region > US
Population > Large

Now let's examine what happens when each column is used in Power BI. When visuals
group on the following columns, here are the results:

City - Power BI displays all the cities by clearing all the default members for City,
State, Country/Region but preserves the default member for Population; Power BI
 cleared the entire ARP for City.

７ Note

Population isn't in the ARP path of City, it's solely related to State and thus
Power BI doesn't clear it.

State - Power BI displays all the States by clearing all default members for City,
State, Country/Region and Population.
Country/Region - Power BI displays all the countries/regions by clearing all default
members for City, State and Country/Region, but preserves the default member for
Population.
City and State - Power BI clears all default members for all columns.

Groups displayed in the visual have their entire ARP path cleared.

If a group isn't displayed in the visual, but is part of the ARP path of another grouped-
on column, the following applies:

Not all branches of the ARP path are cleared automatically.

That group is still filtered by that uncleared default member.

Slicers and filter cards

When you work with slicers or filter cards, the following behavior occurs:

When a slicer or filter card is loaded with data, Power BI groups on the column in
the visual, so the display behavior is the same as described in the previous section.

Since slicers and filter cards are often used to interact with other visuals, the logic of
clearing default members for the affected visuals occurs as explained in the following
table.

For this table, we use the same example data from earlier in this article:
The following rules apply to the way Power BI behaves in these circumstances.

Power BI clears a default member for a specified column, if:

Power BI groups on that column.

Power BI groups on a column related to that column (anywhere in the ARP, up or
down).
Power BI filters on a column that's in the ARP (up or down).
The column has a filter card with ALL stated.
The column has a filter card with any value selected (Power BI receives a filter for
the column).

Power BI doesn't clear a default member for a specified column, if:

The column has a filter card with default stated, and Power BI is grouping on a
column in its ARP.
The column is above another column in the ARP, and Power BI has a filter card for
that other column in default state.

Next steps
This article described the behavior of Power BI when working with default members in
multidimensional models. You might also be interested in the following articles:

Show items with no data in Power BI

Data sources in Power BI Desktop
DirectQuery in Power BI
Article • 11/10/2023

In Power BI Desktop or the Power BI service, you can connect to many different data
sources in different ways. You can import data to Power BI, which is the most common
way to get data. You can also connect directly to some data in its original source
repository, which is called DirectQuery. This article primarily discusses DirectQuery
capabilities.

This article describes:

The different Power BI data connectivity options.

Guidance about when to use DirectQuery rather than import.
Limitations and implications of using DirectQuery.
Recommendations for successfully using DirectQuery.
How to diagnose DirectQuery performance issues.

The article focuses on the DirectQuery workflow when you create a report in Power BI
Desktop, but also covers connecting through DirectQuery in the Power BI service.

７ Note

DirectQuery is also a feature of SQL Server Analysis Services. That feature shares
many details with Direct Query in Power BI, but there are also important
differences. This article primarily covers DirectQuery with Power BI, not SQL Server
Analysis Services.

For more information about using DirectQuery with SQL Server Analysis Services,
see Use DirectQuery for Power BI semantic models and Analysis Services
(preview). You can also download the PDF DirectQuery in SQL Server 2016
Analysis Services .

Power BI data connectivity modes

Power BI connects to a large number of varied data sources, such as:

Online services like Salesforce and Dynamics 365.

Databases like SQL Server, Access, and Amazon Redshift.
Simple files in Excel, JSON, and other formats.
Other data sources like Spark, websites, and Microsoft Exchange.
You can import data from these sources into Power BI. For some sources, you can also
connect using DirectQuery. For a summary of the sources that support DirectQuery, see
Data sources supported by DirectQuery. DirectQuery-enabled sources are primarily
sources that can deliver good interactive query performance.

You should import data into Power BI wherever possible. Importing takes advantage of
the high-performance query engine of Power BI, and provides a highly interactive, fully
featured experience.

If you can't meet your goals by importing data, for example if the data changes
frequently and reports must reflect the latest data, consider using DirectQuery.
DirectQuery is feasible only when the underlying data source can provide interactive
query results in less than five seconds for a typical aggregate query, and can handle the
generated query load. Carefully consider the limitations and implications of using
DirectQuery.

Power BI import and DirectQuery capabilities evolve over time. Changes that provide
more flexibility when using imported data let you import more often, and eliminate
some of the drawbacks of using DirectQuery. Regardless of improvements, the
performance of the underlying data source is a major consideration when using
DirectQuery. If an underlying data source is slow, using DirectQuery for that source
remains unfeasible.

The following sections cover the three options for connecting to data: import,
DirectQuery, and live connection. The remainder of the article focuses on DirectQuery.

Import connections
When you connect to a data source like SQL Server and import data in Power BI
Desktop, the following results occur:

When you initially Get Data, each set of tables you select defines a query that
returns a set of data. You can edit those queries before loading the data, for
example to apply filters, aggregate the data, or join different tables.

Upon load, all the data defined by the queries imports into the Power BI cache.

Building a visual within Power BI Desktop queries the cached data. The Power BI
store ensures the query is fast, and that all changes to the visual reflect
immediately.

Visuals don't reflect changes to the underlying data in the data store. You need to
reimport to refresh the data.
 Publishing the report to the Power BI service as a .pbix file creates and uploads a
semantic model that includes the imported data. You can then schedule data
refresh, for example reimport the data every day. Depending on the location of the
original data source, it might be necessary to configure an on-premises data
gateway for the refresh.

Opening an existing report or authoring a new report in the Power BI service

queries the imported data again, ensuring interactivity.

You can pin visuals or entire report pages as dashboard tiles in the Power BI
service. The tiles automatically refresh whenever the underlying semantic model
refreshes.

DirectQuery connections
When you use DirectQuery to connect to a data source in Power BI Desktop, the
following results occur:

You use Get Data to select the source. For relational sources, you can still select a
set of tables that define a query that logically returns a set of data. For
multidimensional sources like SAP Business Warehouse (SAP BW), you select only
the source.

Upon load, no data is imported into the Power BI store. Instead, when you build a
visual, Power BI Desktop sends queries to the underlying data source to retrieve
the necessary data. The time it takes to refresh the visual depends on the
performance of the underlying data source.

Any changes to the underlying data aren't immediately reflected in existing visuals.
It's still necessary to refresh. Power BI Desktop resends the necessary queries for
each visual, and updates the visual as necessary.

Publishing the report to the Power BI service creates and uploads a semantic
model, the same as for import. However, that semantic model includes no data.

Opening an existing report or authoring a new report in the Power BI service

queries the underlying data source to retrieve the necessary data. Depending upon
the location of the original data source, it might be necessary to configure an on-
premises data gateway to get the data.

You can pin visuals or entire report pages as dashboard tiles. To ensure that
opening a dashboard is fast, the tiles automatically refresh on a schedule, for
 example every hour. You can control refresh frequency depending on how
frequently the data changes and the importance of seeing the latest data.

When you open a dashboard, the tiles reflect the data at the time of the last
refresh, not necessarily the latest changes made to the underlying source. You can
refresh an open dashboard to ensure that it's current.

Live connections
When you connect to SQL Server Analysis Services, you can choose to import the data
or use a live connection to the selected data model. Using a live connection is similar to
DirectQuery. No data is imported, and the underlying data source is queried to refresh
visuals.

For example, when you use import to connect to SQL Server Analysis Services, you
define a query against the external SQL Server Analysis Services source, and import the
data. If you connect live, you don't define a query, and the entire external model shows
in the field list.

This situation also applies when you connect to the following sources, except there's no
option to import the data:

Power BI semantic models, for example connecting to a Power BI semantic model

that's already published to the service, to author a new report over it.

Microsoft Dataverse.

When you publish SQL Server Analysis Services reports that use live connections, the
behavior in the Power BI service is similar to DirectQuery reports in the following ways:

Opening an existing report or authoring a new report in the Power BI service

queries the underlying SQL Server Analysis Services source, possibly requiring an
on-premises data gateway.

Dashboard tiles automatically refresh on a schedule, such as every hour.

A live connection also differs from DirectQuery in several ways. For example, live
connections always pass the identity of the user opening the report to the underlying
SQL Server Analysis Services source.

DirectQuery use cases

Connecting with DirectQuery can be useful in the following scenarios. In several of these
cases, leaving the data in its original source location is necessary or beneficial.

DirectQuery in Power BI offers the greatest benefits in the following scenarios:

The data changes frequently, and you need near real-time reporting.
You need to handle large data without having to pre-aggregate.
The underlying source defines and applies security rules.
Data sovereignty restrictions apply.
The source is a multidimensional source containing measures, such as SAP BW.

Data changes frequently, and you need near real-time

reporting
You can refresh models with imported data at most once per hour, more frequently with
Power BI Pro or Power BI Premium subscriptions. If the data is continually changing, and
it's necessary for reports to show the latest data, using import with scheduled refresh
might not meet your needs. You can stream data directly into Power BI, although there
are limits on the data volumes supported for this case.

Using DirectQuery means that opening or refreshing a report or dashboard always

shows the latest data in the source. The dashboard tiles can also be updated more
frequently, as often as every 15 minutes.

Data is very large

If the data is very large, it's not feasible to import all of it. DirectQuery requires no large
transfer of data, because it queries data in place. However, large data might also make
the performance of queries against that underlying source too slow.

You don't always have to import full detailed data. The Power Query Editor makes it easy
to pre-aggregate data during import. Technically, it's possible to import exactly the
aggregate data you need for each visual. While DirectQuery is the simplest approach to
large data, importing aggregate data might offer a solution if the underlying data
source is too slow for DirectQuery.

These details relate to using Power BI alone. For more information about using large
models in Power BI, see large semantic models in Power BI Premium. There's no
restriction on how frequently the data can be refreshed.

The underlying source defines security rules

When you import data, Power BI connects to the data source by using the current user's
Power BI Desktop credentials, or the credentials configured for scheduled refresh from
the Power BI service. In publishing and sharing reports that have imported data, you
must be careful to share only with users allowed to see the data, or you must define
row-level security as part of the semantic model.

DirectQuery lets a report viewer's credentials pass through to the underlying source,
which applies security rules. DirectQuery supports single sign-on (SSO) to Azure SQL
data sources, and through a data gateway to on-premises SQL servers. For more
information, see Overview of single sign-on (SSO) for gateways in Power BI.

Data sovereignty restrictions apply

Some organizations have policies around data sovereignty, meaning that data can't
leave the organization premises. This data presents issues for solutions based on data
import. With DirectQuery, the data remains in the underlying source location. However,
even with DirectQuery, the Power BI service keeps some caches of data at the visual
level, because of scheduled refresh of tiles.

The underlying data source uses measures

An underlying data source such as SAP HANA or SAP BW contains measures. Measures
mean that imported data is already at a certain level of aggregation, as defined by the
query. A visual that asks for data at a higher-level aggregate, such as TotalSales by Year,
further aggregates the aggregate value. This aggregation is fine for additive measures,
such as Sum and Min, but can be an issue for non-additive measures, such as Average
and DistinctCount.

Easily getting the correct aggregate data needed for a visual directly from the source
requires sending queries per visual, as in DirectQuery. When you connect to SAP BW,
choosing DirectQuery allows this treatment of measures. For more information, see
DirectQuery and SAP BW.

Currently DirectQuery over SAP HANA treats data the same as a relational source, and
produces behavior similar to import. For more information, see DirectQuery and SAP
HANA.

DirectQuery limitations
Using DirectQuery has some potentially negative implications. Some of these limitations
differ slightly depending on the exact source you use. The following sections list general
implications of using DirectQuery, and limitations related to performance, security,
transformations, modeling, and reporting.

General implications
Some general implications and limitations of using DirectQuery follow:

If data changes, you must refresh to show the latest data. Given the use of
caches, there's no guarantee that visuals always show the latest data. For example,
a visual might show transactions in the past day. A slicer change might refresh the
visual to show transactions for the past two days, including recent, newly arrived
transactions. But returning the slicer to its original value could result in it again
showing the cached previous value. Select Refresh to clear any caches and refresh
all the visuals on the page to show the latest data.

If data changes, there's no guarantee of consistency between visuals. Different

visuals, whether on the same page or on different pages, might be refreshed at
different times. If the data in the underlying source is changing, there's no
guarantee that each visual shows the data at the same point in time.

Given that more than one query might be required for a single visual, for example,
to obtain the details and the totals, even consistency within a single visual isn't
guaranteed. To guarantee this consistency would require the overhead of
refreshing all visuals whenever any visual refreshed, along with using costly
features like snapshot isolation in the underlying data source.

You can mitigate this issue to a large extent by selecting Refresh to refresh all of
the visuals on the page. Even for import mode, there's a similar problem of
maintaining consistency when you import data from more than one table.

You must refresh in Power BI Desktop to reflect schema changes. After a report is
published, Refresh in the Power BI service refreshes the visuals in the report. But if
the underlying source schema changes, the Power BI service doesn't automatically
update the available fields list. If tables or columns are removed from the
underlying source, it might result in query failure upon refresh. To update the fields
in the model to reflect the changes, you must open the report in Power BI Desktop
and choose Refresh.

A limit of 1 million rows can return on any query. There's a fixed limit of 1 million
rows that can return in any single query to the underlying source. This limit
generally has no practical implications, and visuals won't display that many points.
However, the limit can occur in cases where Power BI doesn't fully optimize the
queries sent, and requests some intermediate result that exceeds the limit.
 The limit can also occur while building a visual, on the path to a more reasonable
final state. For example, including Customer and TotalSalesQuantity could hit this
limit if there are more than 1 million customers, until you apply some filter. The
error that returns is: The resultset of a query to external data source has
exceeded the maximum allowed size of '1000000' rows.

７ Note

Premium capacities let you exceed the one-million row limit. For more
information, see max intermediate row set count.

You can't change a model from import to DirectQuery mode. You can switch a
model from DirectQuery mode to import mode if you import all the necessary
data. It's not possible to switch back to DirectQuery mode, primarily because of the
feature set that DirectQuery mode doesn't support. For multidimensional sources
like SAP BW, you can't switch from DirectQuery to import mode either, because of
the different treatment of external measures.

Performance and load implications

When you use DirectQuery, the overall experience depends on the performance of the
underlying data source. If refreshing each visual, for example after changing a slicer
value, takes less than five seconds, the experience is reasonable, although might feel
sluggish compared to the immediate response with imported data. If the slowness of
the source causes individual visuals to take longer than tens of seconds to refresh, the
experience becomes unreasonably poor. Queries might even time out.

Along with the performance of the underlying source, the load placed on the source
also impacts performance. Each user who opens a shared report, and each dashboard
tile that refreshes, sends at least one query per visual to the underlying source. The
source must be able to handle such a query load while maintaining reasonable
performance.

Security implications
Unless the underlying data source uses SSO, a DirectQuery report always uses the same
fixed credentials to connect to the source once it's published to the Power BI service.
Immediately after you publish a DirectQuery report, you must configure the credentials
of the user to use. Until you configure the credentials, trying to open the report in the
Power BI service results in an error.
Once you provide the user credentials, Power BI uses those credentials for whoever
opens the report, the same as for imported data. Every user sees the same data, unless
row-level security is defined as part of the report. You must pay the same attention to
sharing the report as for imported data, even if there are security rules defined in the
underlying source.

Connecting to Power BI semantic models and Analysis Services in DirectQuery

mode always uses SSO, so the security is similar to live connections to Analysis
Services.

Alternate credentials aren't supported when making DirectQuery connections to

SQL Server from Power BI Desktop. You can use your current Windows credentials
or database credentials.

You can use multiple data sources in a DirectQuery model by using composite
models. When you use multiple data sources, it's important to understand the
security implications of how data moves back and forth between the underlying
data sources.

Data transformation limitations

DirectQuery limits the data transformations you can apply within Power Query Editor.
With imported data, you can easily apply a sophisticated set of transformations to clean
and reshape the data before using it to create visuals. For example, you can parse JSON
documents, or pivot data from a column to a row form. These transformations are more
limited in DirectQuery.

When you connect to an online analytical processing (OLAP) source like SAP BW, you
can't define any transformations, and the entire external model is taken from the source.
For relational sources like SQL Server, you can still define a set of transformations per
query, but those transformations are limited for performance reasons.

Any transformations must be applied on every query to the underlying source, rather
than once on data refresh. Transformations must be able to reasonably translate into a
single native query. If you use a transformation that's too complex, you get an error that
either it must be deleted or the connection model switched to import.

Also, the Get Data dialog or Power Query Editor use subselects within the queries they
generate and send to retrieve data for a visual. Queries defined in Power Query Editor
must be valid within this context. In particular, it's not possible to use a query with
common table expressions, nor one that invokes stored procedures.
Modeling limitations
The term modeling in this context means the act of refining and enriching raw data as
part of authoring a report using the data. Examples of modeling include:

Defining relationships between tables.

Adding new calculations, like calculated columns and measures.
Renaming and hiding columns and measures.
Defining hierarchies.
Defining column formatting, default summarization, and sort order.
Grouping or clustering values.

You can still make many of these model enrichments when you use DirectQuery, and use
the principle of enriching the raw data to improve later consumption. However, some
modeling capabilities aren't available or are limited with DirectQuery. The limitations are
applied to avoid performance issues.

The following limitations are common to all DirectQuery sources. More limitations might
apply to individual sources.

No built-in date hierarchy: With imported data, every date/datetime column also
has a built-in date hierarchy available by default. For example, if you import a table
of sales orders that includes a column OrderDate, and you use OrderDate in a
visual, you can choose the appropriate date level to use, such as year, month, or
day. This built-in date hierarchy isn't available with DirectQuery. If there's a Date
table available in the underlying source, as is common in many data warehouses,
you can use the Data Analysis Expressions (DAX) time-intelligence functions as
usual.

Date/time support only to the seconds level: For semantic models that use time
columns, Power BI issues queries to the underlying DirectQuery source only up to
the seconds detail level, not milliseconds. Remove milliseconds data from your
source columns.

Limitations in calculated columns: Calculated columns can only be intra-row, that

is they can refer only to values of other columns of the same table, without using
any aggregate functions. Also, the allowed DAX scalar functions, such as LEFT() ,
are limited to those functions that can be pushed to the underlying source. The
functions vary depending upon the exact capabilities of the source. Functions that
aren't supported aren't listed in autocomplete when authoring the DAX query for a
calculated column, and result in an error if used.
 No support for parent-child DAX functions: When in DirectQuery mode, it's not
possible to use the family of DAX PATH() functions that usually handle parent-child
structures, such as charts of accounts or employee hierarchies.

No clustering: When you use DirectQuery, you can't use the clustering capability
to automatically find groups.

Reporting limitations
Almost all reporting capabilities are supported for DirectQuery models. As long as the
underlying source offers a suitable level of performance, you can use the same set of
visualizations as for imported data.

One general limitation is that the maximum length of data in a text column for
DirectQuery semantic models is 32,764 characters. Reporting on longer texts results in
an error.

The following Power BI reporting capabilities can cause performance issues in

DirectQuery-based reports:

Measure filters: Visuals that use measures or aggregates of columns can contain
filters in those measures. For example, the following graphic shows SalesAmount
by Category, but only for categories with more than 20M of sales.

This approach causes two queries to be sent to the underlying source:

The first query retrieves the categories that meet the condition SalesAmount
greater than 20 million.
 The second query retrieves the necessary data for the visual, which includes the
categories that met the WHERE condition.

This approach generally works well if there are hundreds or thousands of

categories, as in this example. Performance can degrade if the number of
categories is much larger. The query fails if there are more than a million
categories.

TopN filters: You can define advanced filters to filter on only the top or bottom N
values ranked by some measure. For example, filters can include the top 10
categories. This approach again sends two queries to the underlying source.
However, the first query returns all categories from the underlying source, and
then the TopN are determined based on the returned results. Depending on the
cardinality of the column involved, this approach can lead to performance issues or
query failures because of the one-million row limit on query results.

Median: Any aggregation, such as Sum or Count Distinct , is pushed to the

underlying source. However, usually the median aggregate isn't supported by the
underlying source. For median , the detail data is retrieved from the underlying
source, and the median is calculated from the returned results. This approach is
reasonable for calculating the median over a relatively small number of results.

Performance issues or query failures can arise if the cardinality is large because of
the one-million row limit. For example, querying for Median Country/Region
Population might be reasonable, but Median Sales Price might not be reasonable.

Advanced text filters like 'contains': Advanced filtering on a text column allows
filters like contains and begins with . These filters can result in degraded
performance for some data sources. In particular, don't use the default contains
filter if you need an exact match. Although the results might be the same
depending on the actual data, the performance might be drastically different
because of indexes.

Multi-select slicers: By default, slicers only allow making a single selection.

Allowing multi-selection in filters can cause performance issues. For example, if the
user selects 10 products of interest, each new selection results in queries being
sent to the source. Although the user can select the next item before the query
completes, this approach results in extra load on the underlying source.

Totals on table visuals: By default, tables and matrices display totals and subtotals.
In many cases, getting the values for such totals requires sending separate queries
to the underlying source. This requirement applies whenever you use
 DistinctCount aggregation, or in all cases that use DirectQuery over SAP BW or

SAP HANA. You can switch off such totals by using the Format pane.

DirectQuery recommendations
This section provides high-level guidance on how to successfully use DirectQuery, given
its implications.

Underlying data source performance

Validate that simple visuals refresh within five seconds, to provide a reasonable
interactive experience. If visuals take longer than 30 seconds to refresh, it's likely that
further issues following report publication will make the solution unworkable.

If queries are slow, examine the queries sent to the underlying source, and the reason
for the slow performance. For more information, see Performance diagnostics.

This article doesn't cover the wide range of database optimization recommendations
across the full set of potential underlying sources. The following standard database
practices apply to most situations:

For better performance, base relationships on integer columns rather than joining
columns of other data types.

Create the appropriate indexes. Index creation generally means using column store
indexes in sources that support them, for example SQL Server.

Update any necessary statistics in the source.

Model design
When you define the model, follow this guidance:

Avoid complex queries in Power Query Editor. Power Query Editor translates a
complex query into a single SQL query. The single query appears in the subselect
of every query sent to that table. If that query is complex, it might result in
performance issues on every query sent. You can get the actual SQL query for a set
of steps by right-clicking the last step under Applied steps in Power Query Editor
and choosing View Native Query.

Keep measures simple. At least initially, limit measures to simple aggregates. If the
measures operate in a satisfactory manner, you can define more complex
measures, but pay attention to performance.
Avoid relationships on calculated columns. In databases where you need to do
multi-column joins, Power BI doesn't allow basing relationships on multiple
columns as the primary key or foreign key. The common workaround is to
concatenate the columns by using a calculated column, and base the join on that
column.

This workaround is reasonable for imported data, but for DirectQuery it results in a
join on an expression. That result usually prevents using any indexes, and leads to
poor performance. The only workaround is to actually materialize the multiple
columns into a single column in the underlying data source.

Avoid relationships on 'uniqueidentifier' columns. Power BI doesn't natively

support a uniqueidentifier datatype. Defining a relationship between
uniqueidentifier columns results in a query with a join that involves a cast. Again,

this approach commonly leads to poor performance. The only workaround is to

materialize columns of an alternative type in the underlying data source.

Hide the 'to' column on relationships. The to column on relationships is

commonly the primary key on the to table. That column should be hidden, but if
hidden, it doesn't appear in the field list and can't be used in visuals. Often the
columns on which relationships are based are actually system columns, for example
surrogate keys in a data warehouse. It's still best to hide such columns.

If the column has meaning, introduce a calculated column that's visible and that
has a simple expression of being equal to the primary key, for example:

SQL

ProductKey_PK (Destination of a relationship, hidden)

ProductKey (= [ProductKey_PK], visible)
ProductName
...

Examine all calculated columns and data type changes. You can use calculated
tables when you use DirectQuery with composite models. These capabilities aren't
necessarily harmful, but they result in queries that contain expressions rather than
simple references to columns. Those queries might result in indexes not being
used.

Avoid bidirectional cross filtering on relationships. Using bidirectional cross

filtering can lead to query statements that don't perform well. For more
information about bidirectional cross filtering, see Enable bidirectional cross-
filtering for DirectQuery in Power BI Desktop, or download the Bidirectional cross-
 filtering white paper. The examples in the paper are for SQL Server Analysis
Services, but the fundamental points also apply to Power BI.

Experiment with setting Assume referential integrity. The Assume referential

integrity setting on relationships enables queries to use INNER JOIN rather than
OUTER JOIN statements. This guidance generally improves query performance,

although it depends on the specifics of the data source.

Don't use the relative data filtering in Power Query Editor. It's possible to define
relative date filtering in Power Query Editor. For example, you can filter to the rows
where the date is in the last 14 days.

However, this filter translates into a filter based on a fixed date, such as the time
the query was authored, as you can see in the native query.

This data is probably not what you want. To ensure the filter is applied based on
the date at the time the report runs, apply the date filter in the report. You can
create a calculated column that calculates the number of days ago by using the
DAX DATE() function, and use that calculated column in the filter.

Report design
When you create a report that uses a DirectQuery connection, follow this guidance:

Consider using query reduction options: Power BI provides report options to send
fewer queries, and to disable certain interactions that cause a poor experience if
the resulting queries take a long time to run. These options apply when you
interact with your report in Power BI Desktop, and also apply when users consume
the report in the Power BI service.

To access these options in Power BI Desktop, go to File > Options and settings >
Options and select Query reduction.

Selections on the Query reduction screen let you show an Apply button for slicers
or filter selections. No queries are sent until you select the Apply button on the
filter or slicer. The queries then use your selections to filter the data. This button
lets you make several slicer and filter selections before you apply them.

Apply filters first: Always apply any applicable filters at the start of building a
visual. For example, rather than drag in TotalSalesAmount and ProductName, and
then filter to a particular year, apply the filter on Year at the beginning.

Each step of building a visual sends a query. Although it's possible to make
another change before the first query completes, this approach still leaves
unnecessary load on the underlying source. Applying filters early generally makes
 those intermediate queries less costly. Failing to apply filters early can result in
hitting the one-million row limit.

Limit the number of visuals on a page: When you open a page or change a page
level slicer or filter, all the visuals on the page refresh. There's a limit on the
number of parallel queries. As the number of visuals increases, some visuals refresh
serially, which increases the time it takes to refresh the page. Therefore, it's best to
limit the number of visuals on a single page, and instead have more, simpler
pages.

Consider switching off interaction between visuals: By default, visualizations on a

report page can be used to cross-filter and cross-highlight the other visualizations
on the page. For example, if you select 1999 on the pie chart, the column chart is
cross-highlighted to show the sales by category for 1999.

Cross-filtering and cross-highlighting in DirectQuery require queries to be

submitted to the underlying source. You should switch off this interaction if the
time taken to respond to users' selections is unreasonably long.

You can use the Query reduction settings to disable cross-highlighting throughout
your report, or on a case-by-case basis. For more information, see How visuals
cross-filter each other in a Power BI report.

Maximum number of connections

You can set the maximum number of connections DirectQuery opens for each
underlying data source, which controls the number of queries concurrently sent to each
data source.

DirectQuery opens a default maximum number of 10 concurrent connections. To change

the maximum number for the current file in Power BI Desktop, go to File > Options and
Settings > Options, and select DirectQuery in the Current File section of the left pane.

The setting is enabled only when there's at least one DirectQuery source in the current
report. The value applies to all DirectQuery sources, and to any new DirectQuery sources
added to that report.

Increasing Maximum connections per data source allows sending more queries, up to
the maximum number specified, to the underlying data source. This approach is useful
when many visuals are on a single page, or many users access a report at the same time.
Once the maximum number of connections is reached, further queries are queued until
a connection becomes available. A higher limit results in more load on the underlying
source, so the setting isn't guaranteed to improve overall performance.

Once you publish a report to the Power BI service, the maximum number of concurrent
queries also depends on fixed limits set on the target environment where the report is
published. Power BI, Power BI Premium, and Power BI Report Server impose different
limits. The table below lists the upper limits of the active connections per data source for
each Power BI environment. These limits apply to cloud data sources and on-premises
data sources such as SQL Server, Oracle, and Teradata.

Environment Upper limit per data source

Power BI Pro 10 active connections

Power BI Premium 30 active connections

Power BI Report Server 10 active connections

７ Note

The maximum number of DirectQuery connections setting applies to all

DirectQuery sources when you enable enhanced metadata, which is the default
setting for all models created in Power BI Desktop.

DirectQuery in the Power BI service

All DirectQuery data sources are supported from Power BI Desktop, and some sources
are also available directly from within the Power BI service. A business user can use
Power BI to connect to their data in Salesforce, for example, and immediately get a
dashboard, without using Power BI Desktop.

Only the following two DirectQuery-enabled sources are available directly in the Power
BI service:

Spark
Azure Synapse Analytics (formerly SQL Data Warehouse)

Even for these two sources, it's still best to start DirectQuery use within Power BI
Desktop. While it's easy to initially make the connection in the Power BI service, there
are limitations on further enhancing the resulting report. For example, in the service it's
not possible to create any calculations, or use many analytical features, or refresh the
metadata to reflect changes to the underlying schema.

The performance of a DirectQuery report in the Power BI service depends on the degree
of load placed on the underlying data source. The load depends on:

The number of users that share the report and dashboard.

The complexity of the report.
Whether the report defines row-level security.
Report behavior in the Power BI service
When you open a report in the Power BI service, all the visuals on the currently visible
page refresh. Each visual requires at least one query to the underlying data source.
Some visuals might require more than one query. For example, a visual might show
aggregate values from two different fact tables, or contain a more complex measure, or
contain totals of a non-additive measure like Count Distinct. Moving to a new page
refreshes those visuals. Refreshing sends a new set of queries to the underlying source.

Every user interaction on the report might result in visuals being refreshed. For example,
selecting a different value on a slicer requires sending a new set of queries to refresh all
of the affected visuals. The same is true for selecting a visual to cross-highlight other
visuals, or changing a filter. Similarly, creating or editing a report requires queries to be
sent for each step on the path to produce the final visual.

There's some caching of results. The refresh of a visual is instantaneous if the exact same
results were recently obtained. If row-level security is defined, these caches aren't shared
across users.

Using DirectQuery imposes some important limitations in some of the capabilities the
Power BI service offers for published reports:

Quick insights aren't supported: Power BI quick insights search different subsets
of your semantic model while applying a set of sophisticated algorithms to
discover potentially interesting insights. Because quick insights require high-
performance queries, this feature isn't available on semantic models that use
DirectQuery.

Using Explore in Excel results in poor performance: You can explore a semantic
model by using the Explore in Excel capability, which lets you create pivot tables
and pivot charts in Excel. This capability is supported for semantic models that use
DirectQuery, but performance is slower than creating visuals in Power BI. If using
Excel is important for your scenarios, account for this issue in deciding whether to
use DirectQuery.

Excel doesn't show hierarchies: For example, when you use Analyze in Excel, Excel
doesn't show any hierarchies defined in Azure Analysis Services models or Power BI
semantic models that use DirectQuery.

Dashboard refresh
In the Power BI service, you can pin individual visuals or entire pages to dashboards as
tiles. Tiles that are based on DirectQuery semantic models refresh automatically by
sending queries to the underlying data sources on a schedule. By default, semantic
models refresh every hour, but you can configure refresh between weekly and every 15
minutes as part of semantic model settings.

If no row-level security is defined in the model, each tile is refreshed once, and the
results are shared across all users. If you use row-level security, each tile requires
separate queries per user to be sent to the underlying source.

There can be a large multiplier effect. A dashboard with 10 tiles, shared with 100 users,
created on a semantic model using DirectQuery with row-level security, results in at
least 1000 queries being sent to the underlying data source for every refresh. Give
careful consideration to the use of row-level security and the configuration of the
refresh schedule.

Query timeouts
A timeout of four minutes applies to individual queries in the Power BI service. Queries
that take longer than four minutes fail. This limit is intended to prevent issues caused by
overly long execution times. You should use DirectQuery only for sources that can
provide interactive query performance.

Performance diagnostics
This section describes how to diagnose performance issues, or how to get more detailed
information to optimize your reports.

Start diagnosing performance issues in Power BI Desktop, rather than in the Power BI
service. Performance issues are often based on the performance of the underlying
source. You can more easily identify and diagnose issues in the more isolated Power BI
Desktop environment.

This approach initially eliminates certain components, such as the Power BI gateway. If
the performance issues don't occur in Power BI Desktop, you can investigate the
specifics of the report in the Power BI service.

The Power BI Desktop Performance analyzer is a useful tool for identifying issues. Try to
isolate any issues to one visual, rather than many visuals on a page. If a single visual on
a Power BI Desktop page is sluggish, use the Performance analyzer to analyze the
queries that Power BI Desktop sends to the underlying source.

You can also view traces and diagnostic information that some underlying data sources
emit. Even if there are no traces from the source, the trace file might contain useful
details of how a query runs and how you can improve it. You can use the following
process to view the queries Power BI sends and their execution times.

Use SQL Server Profiler to see queries

By default, Power BI Desktop logs events during a given session to a trace file called
FlightRecorderCurrent.trc. The trace file is in the Power BI Desktop folder for the current
user, in a folder called AnalysisServicesWorkspaces.

For some DirectQuery sources, this trace file includes all queries sent to the underlying
data source. The following data sources send queries to the log:

SQL Server
Azure SQL Database
Azure Synapse Analytics (formerly SQL Data Warehouse)
Oracle
Teradata
SAP HANA

You can read the trace files by using the SQL Server Profiler, part of the free download
SQL Server Management Studio.

To open the trace file for the current session:

1. During a Power BI Desktop session, select File > Options and settings > Options,
and then select Diagnostics.

2. Under Crash Dump Collection, select Open crash dump/traces folder.

 The Power BI Desktop\Traces folder opens.

3. Navigate to the parent folder and then to the AnalysisServicesWorkspaces folder,

which contains one workspace folder for every open instance of Power BI Desktop.
These folders are named with an integer suffix, such as
AnalysisServicesWorkspace2058279583. The workspace folder is deleted when the
associated Power BI Desktop session ends.

Inside the workspace folder for the current Power BI session, the \Data folder
contains the FlightRecorderCurrent.trc trace file. Make a note of the location.

4. Open SQL Server Profiler, and select File > Open > Trace File.

5. Navigate to or enter the path to the trace file for the current Power BI session, and
open FlightRecorderCurrent.trc.

SQL Server Profiler displays all events from the current session. The following screenshot
highlights a group of events for a query. Each query group has the following events:
 A Query Begin and Query End event, which represent the start and end of a DAX
query generated by changing a visual or filter in the Power BI UI, or from filtering
or transforming data in the Power Query Editor.

One or more pairs of DirectQuery Begin and DirectQuery End events, which
represent queries sent to the underlying data source as part of evaluating the DAX
query.

Multiple DAX queries can run in parallel, so events from different groups can be
interleaved. You can use the ActivityID value to determine which events belong to the
same group.

The following columns are also of interest:

TextData: The textual detail of the event. For Query Begin and Query End events,
the detail is the DAX query. For DirectQuery Begin and DirectQuery End events,
the detail is the SQL query sent to the underlying source. The TextData for the
currently selected event also appears in the pane at the bottom of the screen.
EndTime: The time when the event completed.
Duration: The duration, in milliseconds, it took to run the DAX or SQL query.
Error: Whether an error occurred, in which case the event also displays in red.

To capture a trace to help diagnose a potential performance issue:

1. Open a single Power BI Desktop session, to avoid the confusion of multiple

workspace folders.

2. Do the set of actions of interest in Power BI Desktop. Include a few more actions,
to ensure that the events of interest are flushed into the trace file.
 3. Open SQL Server Profiler and examine the trace. Remember that closing Power BI
Desktop deletes the trace file. Also, further actions in Power BI Desktop don't
immediately appear. You must close and reopen the trace file to see new events.

Keep individual sessions reasonably small, perhaps 10 seconds of actions, not hundreds.
This approach makes it easier to interpret the trace file. There's also a limit on the size of
the trace file. For long sessions, there's a chance of early events being dropped.

Understand the format of queries

The general format of Power BI Desktop queries uses subselects for each table they
reference. The Power Query Editor query defines the subselect queries. For example,
assume you have the following TPC-DS tables in SQL Server:

Running the following query:

SQL

SalesAmount (SUMX(Web_Sales, [ws_sales_price]*[ws_quantity]))

by Item[i_category]
for Date_dim[d_year] = 2000

Results in the following visual in Power BI:

Refreshing that visual produces the SQL query in the following image. There are three
subselect queries for Web_Sales , Item , and Date_dim , which each return all the columns
on the respective table, even though the visual references only four columns.
Power Query Editor defines the exact subselect queries. This use of subselect queries
hasn't been shown to affect performance for the data sources DirectQuery supports.
Data sources like SQL Server optimize away the references to the other columns.

Power BI uses this pattern because the analyst provides the SQL query directly. Power BI
uses the query as provided, without any attempt to rewrite it.

Next steps
For more information about DirectQuery in Power BI, see:

Use DataQuery in Power BI Desktop

This article described aspects of DirectQuery that are common across all data sources.
See the following articles for details about specific sources:

DirectQuery and SAP HANA

DirectQuery and SAP BW
Use DirectQuery for Power BI semantic models and Analysis Services
Live connection and DirectQuery
comparison
Article • 05/13/2024

Live connection is a way of connecting a Power BI report to a published Power BI

semantic model. DirectQuery is a method you can use to connect your semantic model
to data. This article describes the main differences between these concepts.

Live connection
Live connection is a method that lets you build a report in Power BI Desktop without
having to build a semantic model for it. When you create your report in Power BI
Desktop, you can connect it to a semantic model that already exists. A live connection
allows you to rely on existing data, which can be updated without accessing the report.

Using live connection you can connect your report to one of the following data sources:

A semantic model that already exists in Power BI service

An Azure Analysis Services (AAS) database

An on-premises instance of SQL Server Analysis Services (SSAS)

DirectQuery
A Power BI semantic model can have data copied into it during a refresh operation, in
what's called import mode. Or, the semantic model can dynamically request data from a
data source it's connected to using a method called DirectQuery.

When using DirectQuery, your report uses Data Analysis Expression (DAX) queries to get
data. After the semantic model receives the report's DAX query, it generates another set
of queries that are run on your data source, to get the required data. If for example your
data source is an SQL Server database, Power BI will generate SQL queries to get the
data it needs. Other data sources may generate queries in other query languages.

DirectQuery is useful when:

You're working against data sources with a large volume of data

You want to use 'near real-time' data

You can also use DirectQuery with Analysis Services, as described in Using DirectQuery
for Power BI semantic models and Analysis Services.

Related content
For more information, check out the following resources:

Semantic model modes in the Power BI service

Connect to semantic models in the Power BI service from Power BI Desktop

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Connect to data sources in Power BI
Desktop
Article • 01/12/2023

With Power BI Desktop, you can easily connect to the ever expanding world of data. If
you don’t have Power BI Desktop, you can download and install it.

There are all sorts of data sources available in Power BI Desktop. The following image
shows how to connect to data, by selecting Get data > Other > Web.
Example of connecting to data
For this example, we'll connect to a Web data source.

Imagine you’re retiring. You want to live where there’s lots of sunshine, preferable taxes,
and good health care. Or… perhaps you’re a data analyst, and you want that information
to help your customers, as in, help your raincoat manufacturing client target sales where
it rains a lot.

Either way, you find a Web resource that has interesting data about those topics, and
more:

https://www.fool.com/research/best-states-to-retire

Select Get data > Other > Web. In From Web, enter the address.

When you select OK, the Query functionality of Power BI Desktop goes to work. Power
BI Desktop contacts the Web resource, and the Navigator window returns the results of
what it found on that Web page. In this case, it found a table. We're interested in that
table, so we select it from the list. The Navigator window displays a preview.
 

At this point, you can edit the query before loading the table, by selecting Transform
Data from the bottom of the window, or just load the table.

Select Transform Data to load the table and launch Power Query Editor. The Query
Settings pane is displayed. If it's not, select View from the ribbon, then choose Query
Settings to display the Query Settings pane. Here’s what the editor looks like.
 

All those scores are text rather than numbers, and we need them to be numbers. No
problem. Just right-click the column header, and select Change Type > Whole Number
to change them. To choose more than one column, first select a column then choose
Shift, select other adjacent columns, and then right-click a column header to change all
selected columns. Use Ctrl to choose columns that aren't adjacent.

In Query Settings, the APPLIED STEPS reflect any changes that were made. As you make
more changes to the data, Power Query Editor records those changes in the APPLIED
STEPS section, which you can adjust, revisit, rearrange, or delete as necessary.
Other changes to the table can still be made after it's loaded, but for now these changes
are enough. When you're done, select Close & Apply from the Home ribbon, and Power
BI Desktop applies the changes and closes Power Query Editor.

With the data model loaded, in Report view in Power BI Desktop, you can begin creating
visualizations by dragging fields onto the canvas.

Of course, this model is simple, with a single data connection. Most Power BI Desktop
reports have connections to different data sources, shaped to meet your needs, with
relationships that produce a rich data model.
Next steps
There are all sorts of things you can do with Power BI Desktop. For more information on
its capabilities, check out the following resources:

What is Power BI Desktop?

Query overview in Power BI Desktop
Data sources in Power BI Desktop
Shape and combine data in Power BI Desktop
Perform common query tasks in Power BI Desktop

Want to give us feedback? Great! Use the Submit an Idea menu item in Power BI
Desktop or visit Community Feedback . We look forward to hearing from you!
Connect to cloud data sources in the
Power BI service
Article • 05/15/2024

With Power BI, you can share cloud connections for semantic models and paginated
reports, datamarts and dataflows, as well as Power Query Online experiences in Get data,
enabling you to create multiple connection objects to the same cloud data source. For
example, you can create separate connections to the same data source, with different
credentials or privacy settings, and share the connections with others, alleviating the
need for those users to manage their own separate cloud connections.

Types of data connections

The following table shows how various types of connections map to the two primary
connection types: data gateway connections, and direct cloud connections. The new
capability described in this article is Shareable cloud connections.

ﾉ Expand table

Data gateway connections Direct cloud connections

Connections using a personal data gateway Personal cloud connections

Connections using an enterprise or VNET data gateway Shareable cloud connections (new)

Advantages of shareable cloud connections

Connections using a personal cloud connection come with several limitations. For
example, with a personal cloud connection you can only create a single personal cloud
connection object to a given data source. All semantic models that connect to the data
source use the same personal cloud connection object, so if you change the credentials
of the personal cloud connection, all semantic models using that personal cloud
connection are affected. Often that's not a desired outcome.

Another limitation of a personal cloud connection is that they can't be shared with
others, so other users can't bind their semantic models and paginated reports to the
personal cloud connection you own; users must maintain their own personal cloud
connections.
Shareable connections have no such limitations, and provide for more streamlined,
more flexible connection management, including the following:

Support multiple connections to the same data source - support for multiple
connections on the same data source is particularly useful when you want to use
different connection settings for different semantic models, and other artifacts. It's
also useful when you want to assign individual artifacts their own separate
connections, to ensure their connection settings are isolated from each other.

You can share these connections with other users - with shareable connections
you can assign other users Owner permissions, enabling them to manage all
aspects of the connection configuration, including credentials. You can provide
other users with Resharing permissions so they can use and reshare the connection
with others. You can also provide User permissions, enabling them to use the
connection to bind their artifacts to the data source.

Lower the overhead of maintaining data connections and credentials - when

combined with the data source and gateway management experience, you can
centralize data source connection management for gateway and cloud
connections. Such centralization and management is already common for
enterprise and VNET data gateways, for which a gateway administrator creates,
shares, and maintains the connections. With shareable connections, you can now
extend such centralized connection management to cloud data sources as well.

Compare shareable cloud connection to other

connections
By default, when you create a Power BI Desktop report that connects to a cloud data
source, then upload it into a workspace in the Power BI service, Power BI creates a
personal cloud connection and binds it to your semantic model, for which you must
provide credentials. If an existing personal cloud connection is available, you likely
provided the credentials previously.

In contrast, if you have access to at least one shareable cloud connection to the same
data source, you can use the shareable cloud connection, which has already been
configured for you by its owner, instead of having to use your only available personal
cloud connection for the data source.

To use the shareable cloud connection, on the Semantic models settings page, under
Gateway and cloud connections, find Cloud connections and can select the shareable
cloud connection you want to use for the connection, then select Apply. The following
screenshot shows the settings.
Create a new shareable cloud connection
You can create a new shareable cloud connection directly from the Semantic model
settings page. Under Gateway connections > Cloud connections, select the Maps to
dropdown and then select Create a connection.

A pane appears called New connection and automatically populates the configuration
parameters.
Enabling the creation of new connections makes it easy to create separate shareable
cloud connections for individual semantic models, if needed. You can also display the
connection management page from anywhere in the Power BI service by selecting the
Settings gear in the upper right corner of the Power BI service, then select Manage
connections and gateways.

Default connection settings

When connecting to a data source, your Microsoft Entra Single Sign-On (SSO)
credentials are used by default.

You can also use your shareable cloud connection settings instead of your SSO
credentials to connect a semantic model to a data source, and thereby retain the
settings you configured for that shareable cloud connection. This enables you to bind
the data source to the shareable cloud connection, and override the default SSO
connection for that data source.

To select your shareable cloud connection instead of your default SSO settings, select
the shareable cloud connection in the Maps to: drop-down for the data source to which
you want your semantic model to connect, as shown in the following image:

If you don't have a shareable cloud connection, you can select Create a connection and
create a new connection, as described in the previous section of this article.

Using shareable cloud connections with

paginated reports
When you share your paginated report in the Power BI service, you can update the
cloud connections from within the report itself. To modify the cloud connections for
your paginated report, navigate to your workspace in the Power BI service, select the
More button (ellipses) and then select Manage.
Selecting Manage presents a page with several tabs. Select the Reports tab from the
top row, then you can update the connection from within the Cloud connections area,
as shown in the following screenshot.

Limitations and considerations

Shareable cloud connections also share your credentials - when you allow others
to user your shareable cloud connections, it's important to understand that you're
letting others connect their own semantic models, paginated reports, and other
 artifacts to the corresponding data sources by using the connection details and
credentials you provided. Make sure you only share connections (and their
credentials) that you're authorized to share.

You can't mix an Excel on-premises data source with an existing Analysis Services
DirectQuery data source; you can only include an Excel on-premises data source to
your report if it's in a separate query. In such situations, you can map the Excel
data source to a gateway, and leave the Analysis Services DirectQuery cloud data
source as-is.

Related content
For more information about creating shareable cloud connections:

Create and share cloud data sources in the Power BI service

You can do all sorts of things with the Power BI service and Power BI Desktop. For more
information on its capabilities, check out the following resources:

What is Power BI Desktop?

Query overview with Power BI Desktop
Data types in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Create and share cloud data sources in
the Power BI service
Article • 05/03/2024

With Power BI, you can create, share, and manage cloud connections for semantic
models and paginated reports, datamarts, and dataflows, as well as Power Query Online
experiences in Get data, all within the Power BI service user experience.

This article shows you how to create a shareable cloud connection, and then shows you
how to share that connection with others. Creating and sharing shareable cloud
connections have many advantages, as described in advantages of shareable cloud
connections.

Create a shareable cloud connection

To create a shareable cloud connection, go to the Power BI service, select the Settings
gear icon, and from the pane that appears select Manage connections and gateways.

In the window that appears, select New connection and from the pane that appears,
select Cloud.
Enter a name for the new connection, select the appropriate connection type from the
drop-down list, and provide the connection details for your data source. Once you've
filled in the information, select Create.
 

With your connection created, you're ready to share it with others.

７ Note

When a .PBIX file with a cloud data source is published from Power BI Desktop, a
cloud connection is created automatically.

Share a shareable cloud connection

To share a shareable cloud connection that you've already created, go to your
Connections settings in the Power BI service, select the More menu (the ellipses) for the
connection you want to share, and select Manage users.
The Manage users window appears, where you can search users by name or by their
email address, and then grant them the permission level you want them to have. You
must at least grant User permission to allow users to connect their artifacts to the
connection's data source.
Once you've found the user and assigned permission, select Share at the bottom of the
Manage users window to apply your selections.

Assign a shared cloud connection to a semantic

model
Once you've created a shareable cloud connection, you can assign it to a semantic
model.

Open the settings for the semantic model to which you want the shareable connection
to apply, and expand the Gateway and cloud connections section. You'll notice that the
connection is mapped to a Personal Cloud Connection by default.
From the Maps to drop down, select the name of the shareable connection you created
and want to use, then select Apply.

That's it, you've now assigned your shareable cloud connection to the semantic model.

If you haven't created a shareable cloud connection yet when you're using this screen,
you can select the Create a connection option from the drop-down to be taken to the
Manage connections and gateways experience, and all the connection details from the
data source for which you selected the Create a connection drop-down are
prepopulated in the Create new cloud connection form.

Granular access control

Power BI enforces granular access control for shareable cloud connections. Access
control for all data types can be enabled at the tenant, workspace, and semantic model
level. The following image shows how access control can be enforced at the tenant, the
workspace, or the semantic model. Each setting provides granular access control, with
different priority.
If a tenant admin enables granular access control for all connection types, then granular
access control is enforced for the entire organization. Workspace admins and artifact
owners can't overrule granular access control enabled at the tenant level.

If granular access control isn't enforced at the tenant level, workspace admins can
enforce granular access control for their workspaces. And if workspace admins don’t
enforce granular access control, then artifact owners can decide whether to enforce
granular access control for each of their artifacts independently.

By default, granular access control is disabled at all three levels, enabling individual
artifact owners to enforce granular access control for each data connection type
selectively. However, it's likely more efficient to enable granular access control on a
workspace-by-workspace basis.

Related content
For important information about shareable cloud connections, including limitations and
considerations, read the following article:

Connect to cloud data sources in the Power BI service

Feedback
Was this page helpful?  Yes  No
Provide product feedback | Ask the community
Data sources in Power BI Desktop
Article • 05/21/2024

With Power BI Desktop, you can connect to data from many different sources. For a full
list of available data sources, see Power BI data sources.

To see available data sources, in the Home group of the Power BI Desktop ribbon, select
the Get data button label or down arrow to open the Common data sources list. If the
data source you want isn't listed under Common data sources, select More to open the
Get Data dialog box.

Or, open the Get Data dialog box directly by selecting the Get Data icon itself.
 ７ Note

The Power BI team is continually expanding the data sources available to Power BI
Desktop and the Power BI service. As such, you'll often see early versions of work-
in-progress data sources marked as Beta or Preview. Any data source marked as
Beta or Preview has limited support and functionality, and it shouldn't be used in
production environments. Additionally, any data source marked as Beta or Preview
for Power BI Desktop may not be available for use in the Power BI service or other
Microsoft services until the data source becomes generally available (GA).

Data sources
The Get Data dialog box organizes data types in the following categories:

All
File
Database
Microsoft Fabric
Power Platform
Azure
Online Services
Other

The All category includes all data connection types from all categories.

File data sources

The File category provides the following data connections:

Excel Workbook
Text/CSV
XML
JSON
Folder
PDF
Parquet
SharePoint folder

Database data sources

The Database category provides the following data connections:

SQL Server database

Access database
SQL Server Analysis Services database
Oracle database
IBM Db2 database
IBM Informix database (Beta)
IBM Netezza
MySQL database
PostgreSQL database
Sybase database
Teradata database
SAP HANA database
 SAP Business Warehouse Application Server
SAP Business Warehouse Message Server
Amazon Redshift
Impala
Google BigQuery
Google BigQuery (Microsoft Entra ID)(Beta)
Vertica
Snowflake
Essbase
AtScale Models
Actian (Beta)
Amazon Athena
AtScale cubes
BI Connector
Data Virtuality LDW
Denodo
Dremio Software
Dremio Cloud
Exasol
Indexima
InterSystems IRIS (Beta)
Jethro (Beta)
Kyligence
Linkar PICK Style / MultiValue Databases (Beta)
MariaDB
MarkLogic
MongoDB Atlas SQL (Beta)
TIBCO® Data Virtualization
Exact Online Premium (Beta)

７ Note

Some database connectors require that you enable them by selecting File >
Options and settings > Options then selecting Preview Features and enabling the
connector. If you don't see some of the connectors mentioned previously and want
to use them, check your Preview Features settings. Also note that any data source
marked as Beta or Preview has limited support and functionality, and shouldn't be
used in production environments.

Microsoft Fabric
The Microsoft Fabric category provides the following data connections:

Power BI semantic models

Dataflows
Datamarts (preview)
Warehouses
Lakehouses
KQL Database

Power Platform data sources

The Power Platform category provides the following data connections:

Power BI dataflows (Legacy)

Common Data Service (Legacy)
Dataverse
Dataflows

Azure data sources

The Azure category provides the following data connections:

Azure SQL Database

Azure Synapse Analytics SQL
Azure Analysis Services database
Azure Database for PostgreSQL
Azure Blob Storage
Azure Table Storage
Azure Cosmos DB v1
Azure Data Explorer (Kusto)
Azure Data Lake Storage Gen2
Azure HDInsight (HDFS)
Azure HDInsight Spark
HDInsight Interactive Query
Azure Cost Management
Azure HDInsight on AKS Trino (Beta)
Azure Cosmos DB v2 (Beta)
Azure Synapse Analytics workspace (Beta)
Azure Time Series Insights (Beta)
Azure Resource Graph (Beta)
Azure Databricks
Online Services data sources
The Online Services category provides the following data connections:

SharePoint Online List

Microsoft Exchange Online

Dynamics 365 Online (legacy)

Dynamics 365 (Dataverse)

Dynamics NAV

Dynamics 365 Business Central

Dynamics 365 Business Central (on-premises)

Azure DevOps (Boards only)

Azure DevOps Server (Boards only)

Salesforce Objects

Salesforce Reports

Google Analytics

Adobe Analytics

appFigures (Beta)

Data.World - Get Dataset (Beta)

GitHub (Beta)

LinkedIn Sales Navigator (Beta)

Marketo (Beta)

Mixpanel (Beta)

Planview Portfolios

QuickBooks Online (Beta)

Smartsheet

SparkPost (Beta)
SweetIQ (Beta)

Planview Enterprise Architecture

Zendesk (Beta)

Asana (Beta)

Assemble Views

Automation Anywhere

Automy Data Analytics (Beta)

CData Connect Cloud

Dynamics 365 Customer Insights (Beta)

Digital Construction Works Insights

Emigo Data Source

Entersoft Business Suite (Beta)

eWay-CRM

FactSet Analytics

Palantir Foundry

Funnel

Hexagon PPM Smart® API

Industrial App Store

Intune Data Warehouse (Beta)

Planview ProjectPlace

Product Insights (Beta)

Profisee

Quickbase

SoftOne BI (Beta)

Planview IdeaPlace
 TeamDesk (Beta)

Webtrends Analytics (Beta)

Witivio (Beta)

Zoho Creator

Autodesk Construction Cloud

Databricks

Planview OKR (Beta)

Viva Insights

Autodesk Construction Cloud

Other data sources

The Other category provides the following data connections:

Web
SharePoint list
OData Feed
Active Directory
Microsoft Exchange
Hadoop File (HDFS)
Spark
Hive LLAP
R script
Python script
ODBC
OLE DB
Acterys : Model Automation & Planning (Beta)
Amazon OpenSearch Service (Beta)
Anaplan
Solver
Bloomberg Data and Analytics
Cherwell (Beta)
CloudBluePSA (Beta)
Cognite Data Fusion
Delta Sharing
Emplifi Metrics (Beta)
 EQuIS
FactSet RMS (Beta)
FHIR
Google Sheets
Information Grid (Beta)
Jamf Pro (Beta)
Kognitwin
MicroStrategy for Power BI
OpenSearch Project (Beta)
Paxata
QubolePresto (Beta)
Roamler (Beta)
SIS-CC SDMX (Beta)
Shortcuts Business Insights (Beta)
SingleStore Direct Query Connector
Siteimprove
Socialbakers Metrics 1.1.0 (Beta)
SolarWinds Service Desk
Starburst Enterprise
SumTotal
SurveyMonkey
Microsoft Teams Personal Analytics (Beta)
Tenforce (Smart)List
Usercube (Beta)
Vena
Vessel Insight
Wrike (Beta)
Zucchetti HR Infinity (Beta)
BitSight Security Ratings
BQE Core
Celonis EMS
Eduframe (Beta)
Wolters Kluwer CCH Tagetik (Beta)
LinkedIn Learning (Beta)
OneStream (Beta)
Blank Query

７ Note
 At this time, it's not possible to connect to custom data sources secured using
Microsoft Entra ID.

Template apps
You can find template apps for your organization by selecting the Template Apps link
near the bottom of the Get Data window.

Available Template Apps may vary based on your organization.

Connect to a data source

1. To connect to a data source, select the data source from the Get Data window and
select Connect. The following screenshot shows Web selected from the Other data
connection category.
2. A connection window appears. Enter the URL or resource connection information,
and then select OK. The following screenshot shows a URL entered in the From
Web connection dialog box.

3. Depending on the data connection, you might be prompted to provide credentials

or other information. After you provide all required information, Power BI Desktop
connects to the data source and presents the available data sources in the
Navigator dialog box.
 4. Select the tables and other data that you want to load. To load the data, select the
Load button at the bottom of the Navigator pane. To transform or edit the query
in Power Query Editor before loading the data, select the Transform Data button.

Connecting to data sources in Power BI Desktop is that easy. Try connecting to data
from our growing list of data sources, and check back often. We continue to add to this
list all the time.

Use PBIDS files to get data

PBIDS files are Power BI Desktop files that have a specific structure and a PBIDS
extension to identify them as Power BI data source files.

You can create a PBIDS file to streamline the Get Data experience for new or beginner
report creators in your organization. If you create the PBIDS file from existing reports,
it's easier for beginning report authors to build new reports from the same data.

When an author opens a PBIDS file, Power BI Desktop prompts the user for credentials
to authenticate and connect to the data source that the file specifies. The Navigator
dialog box appears, and the user must select the tables from that data source to load
into the model. Users might also need to select the database and connection mode if
none was specified in the PBIDS file.
From that point forward, the user can begin building visualizations or select Recent
Sources to load a new set of tables into the model.

Currently, PBIDS files only support a single data source in one file. Specifying more than
one data source results in an error.

How to create a PBIDS connection file

If you have an existing Power BI Desktop PBIX file already connected to the data you’re
interested in, you can export the connection files from within Power BI Desktop. This
method is recommended, since the PBIDS file can be autogenerated from Desktop. You
can also still edit or manually create the file in a text editor.

1. To create the PBIDS file, select File > Options and settings > Data source settings.

2. In the dialog that appears, select the data source you want to export as a PBIDS
file, and then select Export PBIDS.
 3. In the Save As dialog box, give the file a name, and select Save. Power BI Desktop
generates the PBIDS file, which you can rename and save in your directory, and
share with others.

You can also open the file in a text editor, and modify the file further, including
specifying the mode of connection in the file itself. The following image shows a PBIDS
file open in a text editor.

If you prefer to manually create your PBIDS files in a text editor, you must specify the
required inputs for a single connection and save the file with the PBIDS extension.
Optionally, you can also specify the connection mode as either DirectQuery or Import . If
mode is missing or null in the file, the user who opens the file in Power BI Desktop is

prompted to select DirectQuery or Import.

） Important

Some data sources will generate an error if columns are encrypted in the data
source. For example, if two or more columns in an Azure SQL Database are
encrypted during an Import action, an error will be returned. For more information,
see SQL Database.

PBIDS file examples

This section provides some examples from commonly used data sources. The PBIDS file
type only supports data connections that are also supported in Power BI Desktop, with
the following exceptions: Wiki URLS, Live Connect, and Blank Query.

The PBIDS file doesn't include authentication information and table and schema
information.

The following code snippets show several common examples for PBIDS files, but they
aren't complete or comprehensive. For other data sources, you can refer to the git Data
Source Reference (DSR) format for protocol and address information.

If you're editing or manually creating the connection files, these examples are for
convenience only, aren't meant to be comprehensive, and don't include all supported
connectors in DSR format.

Azure AS

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "analysis-services",
"address": {
"server": "server-here"
},
}
}
]
}
Folder

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "folder",
"address": {
"path": "folder-path-here"
}
}
}
]
}

OData

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "odata",
"address": {
"url": "URL-here"
}
}
}
]
}

SAP BW

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "sap-bw-olap",
"address": {
"server": "server-name-here",
"systemNumber": "system-number-here",
 "clientId": "client-id-here"
},
}
}
]
}

SAP Hana

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "sap-hana-sql",
"address": {
"server": "server-name-here:port-here"
},
}
}
]
}

SharePoint list

The URL must point to the SharePoint site itself, not to a list within the site. Users get a
navigator that allows them to select one or more lists from that site, each of which
becomes a table in the model.

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "sharepoint-list",
"address": {
"url": "URL-here"
},
}
}
]
}
SQL Server

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "tds",
"address": {
"server": "server-name-here",
"database": "db-name-here (optional) "
}
},
"options": {},
"mode": "DirectQuery"
}
]
}

Text file

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "file",
"address": {
"path": "path-here"
}
}
}
]
}

Web

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "http",
 "address": {
"url": "URL-here"
}
}
}
]
}

Dataflow

JSON

{
"version": "0.1",
"connections": [
{
"details": {
"protocol": "powerbi-dataflows",
"address": {
"workspace":"workspace id (Guid)",
"dataflow":"optional dataflow id (Guid)",
"entity":"optional entity name"
}
}
}
]
}

Related content
You can do all sorts of things with Power BI Desktop. For more information on its
capabilities, check out the following resources:

What is Power BI Desktop?

Query overview with Power BI Desktop
Data types in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Dynamic M query parameters in Power
BI Desktop
Article • 11/10/2023

This article describes how to create and work with dynamic M query parameters in
Power BI Desktop. With dynamic M query parameters, model authors can configure the
filter or slicer values that report viewers can use for an M query parameter. Dynamic M
query parameters give model authors more control over the filter selections to
incorporate into DirectQuery source queries.

Model authors understand the intended semantics of their filters, and often know how
to write efficient queries against their data source. With dynamic M query parameters,
model authors can ensure that filter selections incorporate into source queries at the
right point to achieve the intended results with optimum performance. Dynamic M
query parameters can be especially useful for query performance optimization.

Watch Sujata explain and use dynamic M query parameters in the following video, and
then try them out yourself.

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI service.

https://www.microsoft.com/en-us/videoplayer/embed/RE4QLsb?postJsllMsg=true

Prerequisites
To work through these procedures, you must have a valid M query that uses one or
more DirectQuery tables.

Create and use dynamic parameters

The following example passes a single value through to a parameter dynamically.

Add parameters
1. In Power BI Desktop, select Home > Transform data > Transform data to open the
Power Query Editor.
2. In the Power Query Editor, select New Parameters under Manage Parameters in
the ribbon.

3. In the Manage Parameters window, fill out the information about the parameter.
For more information, see Create a parameter.

4. Select New to add more parameters.

 5. When you're done adding parameters, select OK.

Reference the parameters in the M query

1. Once you create the parameters, you can reference them in the M query. To
modify the M query, while you have the query selected, open the Advanced editor.

2. Reference the parameters in the M query, as highlighted in yellow in the following

image:
 

3. When you're done editing the query, select Done.

Create tables of values

Create a table for each parameter with a column that provides the possible values
available to be dynamically set based on filter selection. In this example, you want the
StartTime and EndTime parameters to be dynamic. Since these parameters require a

Date/Time parameter, you generate the possible inputs to dynamically set the date for

the parameter.

1. In the Power BI Desktop ribbon, under Modeling, select New Table.

2. Create a table for the values of the StartTime parameter, for example:

StartDateTable = CALENDAR (DATE(2016,1,1), DATE(2016,12,31))

 3. Create a second table for the values of the EndTime parameter, for example:

EndDateTable = CALENDAR (DATE(2016,1,1), DATE(2016,12,31))

７ Note

Use a column name that's not in an actual table. If you use the same name as
an actual table column, the selected value applies as a filter in the query.

Bind the fields to the parameters

Now that you created the tables with the Date fields, you can bind each field to a
parameter. Binding a field to a parameter means that as the selected field value
changes, the value passes to the parameter and updates the query that references the
parameter.
1. To bind a field, in the Power BI Desktop Model view, select the newly created field,
and in the Properties pane, select Advanced.

７ Note

The column data type should match the M parameter data type.

2. Select the dropdown under Bind to parameter and select the parameter that you
want to bind to the field:
 Since this example is for setting the parameter to a single value, keep Multi-select
set to No, which is the default:

If you set the mapped column to No for Multi-select, you must use a single select
mode in the slicer, or require single select in the filter card.

If your use cases require passing multiple values to a single parameter, set the
control to Yes and make sure your M query is set up to accept multiple values.
Here's an example for RepoNameParameter , which allows multiple values:

3. Repeat these steps if you have other fields to bind to other parameters.
You can now reference this field in a slicer or as a filter:
Enable Select all
In this example, the Power BI Desktop model has a field called Country, which is a list of
countries/regions bound to an M parameter called countryNameMParameter. This
parameter is enabled for Multi-select, but isn't enabled for Select all. To be able to use
the Select all option in a slicer or filter card, take the following added steps:
To enable Select all for Country:

1. In the Advanced properties for Country, enable the Select all toggle, which
enables the Select all value input. Edit the Select all value or note the default
value.

The Select all value passes to the parameter as a list that contains the value you
defined. Therefore, when you define this value or use the default value, make sure
the value is unique and doesn't exist in the field that's bound to the parameter.

2. Launch the Power Query Editor, select the query, and then select Advanced Editor.
Edit the M query to use the Select all value to refer to the Select all option.

3. In the Advanced Editor, add a Boolean expression that evaluates to true if the
parameter is enabled for Multi-select and contains the Select all value, and
 otherwise returns false :

4. Incorporate the result of the Select all Boolean expression into the source query.
The example has a Boolean query parameter in the source query called
includeAllCountries that is set to the result of the Boolean expression from the

previous step. You can use this parameter in a filter clause in the query, such that
false for the Boolean filters to the selected country or region names, and true

effectively applies no filter.

5. Once you update your M query to account for the new Select all value, you can
use the Select all function in slicers or filters.

For reference, here's the full query for the preceding example:

Kusto

let
selectedcountryNames = if Type.Is(Value.Type(countryNameMParameter),
List.Type) then
Text.Combine({"'", Text.Combine(countryNameMParameter, "','") , "'"})
 else
Text.Combine({"'" , countryNameMParameter , "'"}),

selectAllCountries = if Type.Is(Value.Type(countryNameMParameter),
List.Type) then
List.Contains(countryNameMParameter, "__SelectAll__")
else
false,

KustoParametersDeclareQuery = Text.Combine({"declare query_parameters(",

"startTimep:datetime = datetime(",
DateTime.ToText(StartTimeMParameter, "yyyy-MM-dd hh:mm"), "), " ,
"endTimep:datetime = datetime(",
DateTime.ToText(EndTimeMParameter, "yyyy-MM-dd hh:mm:ss"), "), ",
"includeAllCountries: bool = ",
Logical.ToText(selectAllCountries) ,",",
"countryNames: dynamic = dynamic([",
selectedcountryNames, "]));" }),

ActualQueryWithKustoParameters =
"Covid19
| where includeAllCountries or Country
in(countryNames)
| where Timestamp > startTimep and Timestamp
< endTimep
| summarize sum(Confirmed) by Country,
bin(Timestamp, 30d)",

finalQuery = Text.Combine({KustoParametersDeclareQuery,
ActualQueryWithKustoParameters}),

Source = AzureDataExplorer.Contents("help", "samples", finalQuery,

[MaxRows=null, MaxSize=null, NoTruncate=null,
AdditionalSetStatements=null]),
#"Renamed Columns" = Table.RenameColumns(Source,{{"Timestamp", "Date"},
{"sum_Confirmed", "Confirmed Cases"}})
in
#"Renamed Columns"

Potential security risk

Report readers who can dynamically set the values for M query parameters may be able
to access more data or trigger modifications to the source system by using injection
attacks. This possibility depends on how you reference the parameters in the M query
and what values you pass to the parameters.

For example, you have a parameterized Kusto query constructed as follows:

Kusto
 Products
| where Category == [Parameter inserted here] & HasReleased == 'True'
| project ReleaseDate, Name, Category, Region

There are no issues with a friendly user who passes an appropriate value for the
parameter, for example, Games :

| where Category == 'Games' & HasReleased == 'True'

However, an attacker may be able to pass a value that modifies the query to get access
to more data, for example, 'Games'// :

Products
| where Category == 'Games'// & HasReleased == 'True'
| project ReleaseDate, Name, Category, Region

In this example, the attacker can get access to information about games that haven't
released yet by changing part of the query into a comment.

Mitigate the risk

To mitigate the security risk, avoid string concatenation of M parameter values within
the query. Instead, consume those parameter values in M operations that fold to the
source query, so that the M engine and connector construct the final query.

If a data source supports importing stored procedures, consider storing your query logic
there and invoking it in the M query. Alternatively, if available, use a parameter passing
mechanism that's built in to the source query language and connectors. For example,
Azure Data Explorer has built-in query parameter capabilities that are designed to
protect against injection attacks.

Here are some examples of these mitigations:

Example that uses the M query's filtering operations:

Kusto

Table.SelectRows(Source, (r) => r[Columns] = Parameter)

 Example that declares the parameter in the source query, or passes the parameter
value as an input to a source query function:

Kusto

declare query_parameters (Name of Parameter : Type of Parameter);

Example of directly calling a stored procedure:

Kusto

let CustomerByProductFn = AzureDataExplorer.Contents("Help",

"ContosoSales"){[Name="CustomerByProduct"]}[Data] in
CustomerByProductFn({1, 3, 5})

Considerations and limitations

There are some considerations and limitations when you use dynamic M query
parameters:

A single parameter can't be bound to multiple fields nor vice-versa.

Dynamic M query parameters don't support aggregations.
Dynamic M query parameters don't support row-level security (RLS).
Parameter names can't be Data Analysis Expressions (DAX) reserved words nor
contain spaces. You can append Parameter to the end of the parameter name to
help avoid this limitation.
Table names can't contain spaces or special characters.
If your parameter is the Date/Time data type, you need to cast it within the M
query as DateTime.Date(<YourDateParameter>) .
If you use SQL sources, you might get a confirmation dialog every time the
parameter value changes. This dialog is due to a security setting: Require user
approval for new native database queries. You can find and turn off this setting in
the Security section of the Power BI Desktop Options.
Dynamic M query parameters may not work when accessing a semantic model in
Excel.
Dynamic M query parameters are not supported on Power BI Report Server.

Unsupported out-of-box parameter types

Any
Duration
 True/False
Binary

Unsupported filters
Relative time slicer or filter
Relative date
Hierarchy slicer
Multifield include filter
Exclude filters / Not filters
Cross-highlighting
Drilldown filter
Cross drill filter
Top N filter

Unsupported operations
And
Contains
Less than
Greater than
Starts with
Does not start with
Is not
Does not contain
Is blank
Is not blank

Next steps
For more information about Power BI Desktop capabilities, check out the following
resources:

DirectQuery in Power BI Desktop

What is Power BI Desktop?
Query overview with Power BI Desktop
Data types in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop
Create a Power BI semantic model
directly from Log Analytics
Article • 11/10/2023

You can quickly create a Power BI semantic model directly from a Log Analytics query.
The semantic model will be full-fledged Power BI semantic model that you can use to
create reports, analyze in Excel, and more.

Creating a semantic model directly from a Log Analytics query is an easy and quick way
to share a semantic model, because if you save it to a shared workspace, everyone with
the sufficient permissions in the workspace can use it. You can also use semantic model
sharing to share it with other users who don’t have a role in the workspace.

This feature creates a semantic model in the Power BI service directly from a Log
Analytics query. If you need to model or transform the data in ways that aren't available
in the service, you can also export the query from Log Analytics, paste it into Power BI
Desktop, and do your advanced modeling there. For more information, see Create
Power BI semantic models and reports from Log Analytics queries.

Prerequisites
You must have a Power BI account to be able to use this functionality.

Create a dataset from Log Analytics

To create a Power BI dataset from a Log Analytics query:

1. Open and run the Log Analytics query you want to use to create the Power BI
database.

2. In the actions bar, select Export > Export to Power BI.

3. Power BI will open and a dialog will ask you to name the semantic model and
choose a workspace to save it in. By default the semantic model will be given the
same name as the query and saved to My workspace. You can choose your own
name and destination workspace. If you're a free user in Power BI, you'll only be
able to save to My workspace.

The dialog also shows the URL of the Log Analytics data source. To prevent
inadvertently exposing sensitive data, make sure that you recognize the data
source and are familiar with the data. Select Review data if you want to check the
Log Analytic query results before allowing export to continue. For more
information about when reviewing the data might be a good idea, see Reviewing
the Log Analytics data.

4. Select Continue. Your semantic model will be created, and you'll be taken to the
details page of the new semantic model. From there you can do all the things you
can do with a regular Power BI semantic model - refresh the data, share the
semantic model, create new reports, and more. See semantic model details for
more information.

７ Note

If you've connected to Log Analytics from Power BI before, you'll be asked to

choose which credentials to use to for the connection between Power BI and
Log Analytics before being taken to the semantic model details page. For help
 deciding which credentials to choose, see Choosing which credentials to
authenticate with.

To keep the data fresh after you've created the semantic model, either refresh the data
manually or set up scheduled refresh.

Reviewing the Log Analytics data

When you export data from a Log Analytics query to Power BI, a redirect URL is created
that includes all the parameters needed to launch the create semantic model process in
Power BI. If you're the person who selected Export to BI in Log Analytics, you probably
don't need to worry about reviewing the data because you most likely are familiar with
the data you're exporting.

Reviewing the data is important if you weren't the one who exported the Log Analytics
data, but rather received a link from someone for creating a semantic model from Log
Analytics. In such a case, you might not be familiar with the data that is being exported,
and hence it's important to review it to make sure that no sensitive data is inadvertently
being exposed.

Choosing which credentials to authenticate

with
When you export data from Log Analytics to Power BI, Power BI connects to Log
Analytics to get the data. In order to connect, it needs to authenticate with Log
Analytics.
If you get the following dialog, it means that you've already established a connection to
Log Analytics in the past. The credentials you used at that time may or may not be
different than the credentials of your current sign in. You need to choose whether to
continue using the sign-in details you used the last time you connected (The credentials
I used to connect to Power BI last time), or whether the connection should use your
current sign-in credentials from now on (My current credentials (these may be the same
or different)).

Why is this important?

The Power BI view of the Log Analytics data is determined by the permissions of the
account used to establish the Power BI connection to the Log Analytics data source.

If you let Power BI use the sign-in details you used last time for the connection, the data
you'll see in the semantic model you're creating may differ from what you see in Log
Analytics. This is because the data that is shown in the semantic model is what the
account with the credentials you used last time can see in Log Analytics.

If you replace the credentials you used last time with your current sign-in credentials,
the data you see in the semantic model you're creating will be exactly the same as what
you see in Log Analytics. However, since the connection now uses your current login
credentials, views of the data in semantic models you might have created previously
from that Log Analytics query might also change, and this could affect reports and
other downstream items that users might have created based on those semantic
models.

Take the above considerations into account when you make your choice.
If you've never previously connected to Log Analytics from Power BI, Power BI will
automatically use your current credentials to establish the connection, and you won't
see this dialog.

Considerations and limitations

This flow does not support business-to-business (B2B) scenarios or scenarios
where authentication takes place against a service principal.
If the Windows Azure Service Management API, the Log Analytics API service, or
both, are configured to use multi-factor authentication, then in order for this flow
to work, Power BI must also be configured to use multi-factor authentication.
Consult your organization's IT support if you encounter a problem related to this
consideration.

Next steps
Log Analytics integration with Power BI
Semantic model details
Share access to a semantic model
Create a Power BI semantic model
directly from a SharePoint list
Article • 11/10/2023

You can quickly create a Power BI semantic model directly from a SharePoint list. The
semantic model will be full-fledged Power BI semantic model that you can use to create
reports, analyze in Excel, and more.

Creating a semantic model directly from a SharePoint list is an easy and quick way to
share a semantic model, because if you save it to a shared workspace, everyone with the
sufficient permissions in the workspace can use it. You can also use semantic model
sharing to share it with other users who don’t have a role in the workspace.

To keep the data fresh after you've created the semantic model, either refresh the data
manually or set up scheduled refresh.

This feature creates a semantic model in the Power BI service directly from a SharePoint
list. If you need to model or transform the data in ways that aren't available in the
service, you can also connect to the SharePoint list from Power BI Desktop. For more
information, see Create a report on a SharePoint List in Power BI Desktop.

Prerequisites
You must have a Power BI account to be able to use this functionality.

Create a semantic model from a SharePoint list

To create a Power BI semantic model from a SharePoint list:

1. Open your SharePoint list.

2. In the actions bar, select Export > Export to Power BI.

3. Power BI will open and a dialog will ask you to name the semantic model and
choose a workspace to save it in. By default the semantic model will be given the
same name as the SharePoint list and saved to My workspace. You can choose your
own name and destination workspace. If you're a free user in Power BI, you'll only
be able to save to My workspace.

The dialog also shows the URL of the data source (SharePoint site) and name of
the SharePoint list. To prevent inadvertently exposing sensitive data, make sure
that you recognize the data source and are familiar with the data. Select Review
data if you want to check the SharePoint list before allowing export to continue.
For more information about when reviewing the data might be a good idea, see
Reviewing the SharePoint list data.

4. Select Continue. Your semantic model will be created, and you'll be taken to the
details page of the new semantic model. From there you can do all the things you
can do with a regular Power BI semantic model - refresh the data, share the
 semantic model, create new reports, and more. See semantic model details for
more information.

７ Note

If you've connected to the SharePoint site from Power BI before, you'll be

asked to choose which credentials to use to for the connection between
Power BI and the Sharepoint site before being taken to the semantic model
details page. For help deciding which credentials to choose, see Choosing
which credentials to authenticate with.

To keep the data fresh after you've created the semantic model, either refresh the data
manually or set up scheduled refresh.

Reviewing the SharePoint list data

When you export a SharePoint list to Power BI, a redirect URL is created that includes all
the parameters needed to launch the create semantic model process in Power BI. If
you're the person who selected Export to BI in your SharePoint list, you probably don't
need to worry about reviewing the data because you most likely are familiar with the
data you're exporting.

Reviewing the data is important if you weren't the one who exported the SharePoint list,
but rather received a link from someone for creating a semantic model from a
SharePoint list. In such a case, you might not be familiar with the data that is being
exported, and hence it's important to review it to make sure that no sensitive data is
inadvertently being exposed.
Choosing which credentials to authenticate
with
When you export a SharePoint list to Power BI, Power BI connects to the SharePoint site
to get the data from the list. In order to connect, it needs to authenticate with
SharePoint.

If you get the following dialog, it means that you've already established a connection to
the SharePoint site in the past. The credentials you used at that time may or may not be
different than the credentials of your current sign in. You need to choose whether to
continue using the sign-in details you used the last time you connected (The credentials
I used to connect to Power BI last time), or whether the connection should use your
current sign-in credentials from now on (My current credentials (these may be the same
or different)).

Why is this important?

The Power BI view of the SharePoint list data is determined by the permissions of the
account used to establish the Power BI connection to the SharePoint data source (that is,
the SharePoint site).

If you let Power BI use the sign-in details you used last time for the connection, the data
you'll see in the semantic model you're creating may differ from what you see in the
SharePoint list. This is because the data that is shown in the semantic model is what the
account with the credentials you used last time can see in the SharePoint list.

If you replace the credentials you used last time with your current sign-in credentials,
the data you see in the semantic model you're creating will be exactly the same as what
you see in the SharePoint list. However, since the connection now uses your current
login credentials, views of the data in semantic models you might have created
previously from that SharePoint site might also change, and this could affect reports
and other downstream items that users might have created based on those semantic
models.

Take the above considerations into account when you make your choice.

If you've never previously connected to the SharePoint site from Power BI, Power BI will
automatically use your current credentials to establish the connection, and you won't
see this dialog.

Considerations and limitations

The semantic model won't be created if the SharePoint list contains values with
more than four digits after a decimal place (".")
The sensitivity label (if any) of the SharePoint list isn't inherited by the semantic
model that is created.
This flow does not support business-to-business (B2B) scenarios or scenarios
where authentication takes place against a service principal.
If the SharePoint service is configured to use multi-factor authentication, then in
order for this flow to work, Power BI must also be configured to use multi-factor
authentication. Consult your organization's IT support if you encounter a problem
related to this consideration.

Next steps
Semantic model details
Share access to a semantic model
Create a report on a SharePoint List in
Power BI Desktop
Article • 11/10/2023

Many teams and organizations use lists in SharePoint Online to store data because it's
easy to set up and easy for users to update. Sometimes a chart is a much easier way for
users to quickly understand the data rather than looking at the list itself. In this tutorial,
you learn how to transform your SharePoint list data into a Power BI report.

Watch this five-minute tutorial video, or scroll down for step-by-step instructions.

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI service.

https://www.youtube-nocookie.com/embed/OZO3x2NF8Ak

In the Power BI service, you can also create a report quickly from data in a SharePoint
list.

If your purpose is to quickly create a semantic model in the Power BI service, you can do
so directly from the SharePoint list. For more information, see Create a semantic model
from a SharePoint list.

Part 1: Connect to your SharePoint List

1. If you don't have it already, download and install Power BI Desktop .

2. Open Power BI Desktop and in the Home tab of the ribbon, select Get data >
More.

3. Select Online Services, and then select SharePoint Online List.

4. Select Connect.

5. Find the address (also known as a URL) of your SharePoint Online site that contains
your list. From a page in SharePoint Online, you can usually get the site address by
selecting Home in the navigation pane, or the icon for the site at the top, then
copying the address from your web browser's address bar.

Watch a video of this step:

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI
service.

https://www.youtube-nocookie.com/embed/OZO3x2NF8Ak?start=48&end=90
 6. In Power BI Desktop, paste the address into the Site URL field of the SharePoint
Online Lists dialog box, and then select OK.

7. You might or might not see a SharePoint access screen like the following image. If
you don't see it, skip to step 10. If you do see it, select Microsoft Account on the
left side of the page.

8. Select Sign In, and enter the user name and password you use to sign in to
Microsoft 365.

9. When you finish signing in, select Connect.

10. On the left side of the Navigator dialog box, select the checkbox beside the
SharePoint list you want to connect to.
 11. Select Load. Power BI loads your list data into a new report.

Part 2: Create a report

1. On the left side of the Power BI Desktop screen, select the Data icon to see that
your SharePoint list data was loaded.

2. Make sure your list columns with numbers show the Sum, or Sigma, icon in the
Data pane on the right. For any that don't, select the column header in the table
view, select the Structure group in the Column tools tab, then change the Data
type to Decimal Number or Whole Number, depending on the data. If prompted
to confirm your change, select Yes. If your number is a special format, like currency,
you can also choose that by setting the Format in the Formatting group.

Watch a video of this step:

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI
service.

https://www.youtube-nocookie.com/embed/OZO3x2NF8Ak?start=147&end=204
 3. On the left side of the Power BI Desktop screen, select the Report icon.

4. Select columns you want to visualize by selecting the checkboxes beside them in
the Fields pane on the right.

Watch a video of this step:

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI
service.

https://www.youtube-nocookie.com/embed/OZO3x2NF8Ak?start=215&end=252

5. Change the visual type if you need to.

6. You can create multiple visualizations in the same report by deselecting the
existing visual, then selecting checkboxes for other columns in the Fields pane.

7. Select Save to save your report.

Next steps
Create a report quickly from a SharePoint list
Connect to LinkedIn Sales Navigator in
Power BI Desktop
Article • 03/20/2023

In Power BI Desktop, you can connect to LinkedIn Sales Navigator to help find and build
relationships just like any other data source in Power BI Desktop, and create ready-made
reports about your progress.

To connect to LinkedIn data using the LinkedIn Sales Navigator, you need to have a
LinkedIn Sales Navigator Enterprise plan, and either be an Admin or Reporting User on
the Sales Navigator Contract.

The following video provides a quick tour and tutorial for using the LinkedIn Sales
Navigator template app, which is described in detail later in this article.

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI service.

https://www.youtube-nocookie.com/embed/ZqhmaiORLw0

Connect to LinkedIn Sales Navigator

To connect to LinkedIn Sales Navigator data, follow the instructions in the Power Query
LinkedIn Sales Navigator article.
Using the LinkedIn Sales Navigator template
app
To make using the LinkedIn Sales Navigator as easy as possible, you can use the
template app that automatically creates a ready-made report from your LinkedIn Sales
Navigator data.

When you download the app, you can select whether to connect to your data, or
explore the app with sample data. You can always go back and connect to your own
LinkedIn Sales Navigator data after you explore the sample data.

You can get the LinkedIn Sales Navigator template app from the following link:

LinkedIn Sales Navigator template app

The template app provides four tabs to help analyze and share your information:

Usage
Search
InMail
SSI

The Usage tab shows your overall LinkedIn Sales Navigator data.

The Search tab lets you drill deeper into your search results.

The InMail tab provides insights into your InMail usage, including number of InMails
sent, acceptance rates, and other useful information.
 

The SSI tab provides more details into your social selling index (SSI).

To go from the sample data to your own data, select edit app in the top-right corner
(the pencil icon) and then select Connect your data from the screen that appears.

From there you can connect your own data, selecting how many days of data to load.
You can load up to 365 days of data. You need to sign in, again using the same email
address you use to sign in to LinkedIn Sales Navigator through the website.
The template app then refreshes the data in the app with your data. You can also set up
a scheduled refresh, so the data in your app is as current as your refresh frequency
specifies.

Once the data updates, you can see the app populated with your own data.

Getting help
If you run into problems when connecting to your data, you can contact LinkedIn Sales
Navigator support .

Next steps
There are all sorts of data you can connect to using Power BI Desktop. For more
information on data sources, check out the following resources:

What is Power BI Desktop?

Data Sources in Power BI Desktop
Shape and Combine Data with Power BI Desktop
Connect to Excel workbooks in Power BI Desktop
Enter data directly into Power BI Desktop
Connect to semantic models in the
Power BI service from Power BI Desktop
Article • 05/22/2024

In Power BI Desktop, you can create a data model and publish it to the Power BI service.
Then you and others can establish a live connection to the shared semantic model that's
in the Power BI service, and create many different reports from that common data
model. You can use the Power BI service live connection feature to create multiple
reports in .pbix files from the same semantic model, and save them to different
workspaces.

This article discusses the benefits, best practices, considerations, and limitations of the
Power BI service live connection feature.

Power BI live connection and report lifecycle

management
One challenge with the popularity of Power BI is the resulting proliferation of reports,
dashboards, and underlying data models. It's easy to create compelling reports in Power
BI Desktop, publish those reports in the Power BI service, and create great dashboards
from those semantic models.

Because report creators often use the same or nearly the same semantic models,
knowing which semantic model a report is based on and the freshness of that semantic
model becomes a challenge. The Power BI service live connection addresses that
challenge by using common semantic models to make it easier and more consistent to
create, share, and expand on reports and dashboards.

Create and share a semantic model everyone can use

A business analyst on your team who is skilled at creating good data models, also called
semantic models, can create a semantic model and report and then share that report in
the Power BI service.

If everyone on the team created their own versions of the semantic model and shared
their reports with the team, there would be many reports from different semantic
models in your team's Power BI workspace. It would be hard to tell which report was the
most recent, whether the semantic models were the same, or what the differences were.

With the Power BI service live connection feature, other team members can use the
analyst's published semantic model for their own reports in their own workspaces.
Everyone can use the same solid, vetted, published semantic model to build their own
unique reports.

Connect to the semantic model by using a Power BI

service live connection
In Power BI Desktop, the team business analyst creates a report and the semantic model
the report is based on. The analyst then publishes the report to the Power BI service, and
the report shows up in the team's workspace. For more information about workspaces,
see Workspaces in Power BI.

The business analyst can use the Build permission setting to make the report available
for anyone in or out of the workspace to see and use. Team members in and out of the
team workspace can now establish a live connection to the shared data model by using
the Power BI service live connection feature. Team members can create their own unique
reports, from the original semantic model, in their own workspaces.

The following image shows how one Power BI Desktop report and its data model
publish to the Power BI service. Others users connect to the data model by using the
Power BI service live connection, and base their own unique reports in their own
workspaces on the shared semantic model.

Set up and use a Power BI service live

connection
You can see the usefulness of the Power BI service live connection for report lifecycle
management. Now find out how to get from a great report and semantic model to a
shared semantic model that teammates can use in Power BI.

Publish a Power BI report and semantic model

The first step in using a Power BI service live connection to manage report lifecycle is to
publish a report and semantic model for teammates to use.
1. To publish the report, from Power BI Desktop, select Publish from the Home tab.

If you're not signed in to the Power BI service account, Power BI prompts you to
sign in.

2. Select the workspace destination to publish the report and semantic model to, and
choose Select. Anyone who has Build permission can then access that semantic
model. You can set Build permission in the Power BI service after publishing.
The publishing process begins, and Power BI Desktop shows the progress.

Once complete, Power BI Desktop shows success, and provides links to the report
in the Power BI service and to quick insights about the report.
 3. Now that your report with its semantic model is in the Power BI service, you can
promote it, or attest to its quality and reliability. You can also request for the report
to be certified by a central authority in your Power BI tenant. For more information,
see Endorse your content.

4. The last step is to set Build permission in the Power BI service for the semantic
model the report is based on. Build permission determines who can see and use
your semantic model. You can set Build permission in the workspace itself, or when
you share an app from the workspace. For more information, see Build permission
for shared semantic models.

Establish a Power BI service live connection to the

published semantic model
Teammates who have access to the workspace where the report and semantic model
were published can connect to the semantic model and build their own reports. To
establish a connection to a published report and create your own report based on the
published semantic model:

1. In Power BI Desktop, on the Home tab, select Get data > Power BI semantic
models.

Or, select Get data, and on the Get Data screen, select Power Platform in the left
pane, select Power BI semantic models, and then select Connect.

If you're not signed in, Power BI prompts you to sign in.

2. The Data hub shows the workspaces you're a member of, and all the shared
semantic models you have Build permission for in any workspace.

To find the semantic model you want, you can:

Filter the list to My data or semantic models that are Endorsed in your org.
Search for a specific semantic model or filter by keyword.
See semantic model name, owner, workspace, last and next refresh time, and
sensitivity.
 3. Select a semantic model, and then select Connect to establish a live connection to
the selected semantic model. Power BI Desktop loads the semantic model fields
and their values in real time.

Now you and others can create and share custom reports, all from the same semantic
model. This approach is a great way to have one knowledgeable person create a well-
formed semantic model. Many teammates can use that shared semantic model to create
their own reports.
Considerations and limitations
When you use the Power BI service live connection, keep a few considerations and
limitations in mind.

Only users with Build permission for a semantic model can connect to a published
semantic model by using the Power BI service live connection.
Hidden columns will become visible to users with Build permissions when they
create live connections to the semantic model in Power BI Desktop.
Free users only see datasets that are in their My Workspace and in Premium or
Fabric based workspaces.
Because this connection is live, left navigation and modeling are disabled. The
behavior is similar to a SQL Server Analysis Services (SSAS) connection. However,
composite models in Power BI make it possible to combine data from different
sources. For more information, see Use composite models in Power BI Desktop.
Because this connection is live, row-level security (RLS) and similar connection
behaviors are enforced. This behavior is the same as when connected to SSAS.
If the owner modifies the original shared .pbix file, the shared semantic model and
report in the Power BI service are overwritten. Reports based on the semantic
model aren't overwritten, but any changes to the semantic model reflect in the
report.
Members of a workspace can't replace the original shared report. If they try to do
so, they get a prompt to rename the file and publish.
If Members are required to publish, they need to download using the option A
Copy of your report and data. Make the necessary changes then publish the report.
If you delete the shared semantic model in the Power BI service, reports based on
that semantic model will no longer work properly or display visuals. You can no
longer access that semantic model from Power BI Desktop.
Reports that share a semantic model on the Power BI service don't support
automated deployments that use the Power BI REST API.
Since the Power BI service connection is live, connecting to a dataset with shared
report in other users' My Workspace is not supported.

Related content
For more information on DirectQuery and other Power BI data connection features,
check out the following resources:

Use DirectQuery in Power BI

Data sources supported by DirectQuery
 Using DirectQuery for Power BI semantic models and Azure Analysis Services
(preview)

For more information about Power BI, see the following articles:

What is Power BI Desktop?

Query overview with Power BI Desktop
Data types in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop
Publish semantic models and reports from Power BI Desktop

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Import Excel workbooks into Power BI
Desktop
Article • 01/09/2023

With Power BI Desktop, you can easily import Excel workbooks that contain Power
Query queries and Power Pivot models into Power BI Desktop. Power BI Desktop
automatically creates reports and visualizations based on the Excel workbook. Once
imported, you can continue to improve and refine those reports with Power BI Desktop,
using the existing features and new features released with each Power BI Desktop
monthly update.

Import an Excel workbook

1. To import an Excel workbook into Power BI Desktop, select File > Import > Power
Query, Power Pivot, Power View.

2. From the Open window, select an Excel workbook to import.

Although there's currently no limitation on the size or number of objects in the

workbook, larger workbooks take longer for Power BI Desktop to analyze and
 import.

７ Note

To load or import Excel files from shared OneDrive for work or school folders
or from Microsoft 365 group folders, use the URL of the Excel file, and input it
into the Web data source in Power BI Desktop. There are a few steps you need
to follow to properly format the OneDrive for work or school URL; for
information and the correct series of steps, see Use OneDrive for work or
school links in Power BI Desktop.

3. From the import dialog box that appears, select Start.

Power BI Desktop analyzes the workbook and converts it into a Power BI Desktop
file (.pbix). This action is a one-time event. Once created with these steps, the
Power BI Desktop file has no dependence on the original Excel workbook. You can
modify, save, and share it without affecting the original workbook.

After the import finishes, a summary page appears that describes the items that
were converted. The summary page also lists any items that couldn't be imported.
4. Select Close.

Power BI Desktop imports the Excel workbook and loads a report based on the
workbook contents.
After the workbook is imported, you can continue working on the report. You can create
new visualizations, add data, or create new report pages by using any of the features
and capabilities included in Power BI Desktop.

Which workbook elements import?

Power BI Desktop can import the following elements, commonly referred to as objects, in
Excel.

Object in Final result in Power BI Desktop file

Excel
workbook

Power Query All Power Query queries from Excel are converted to queries in Power BI Desktop.
queries If there are query groups defined in the Excel Workbook, the same organization
replicates in Power BI Desktop. All queries are loaded unless they're set to Only
Create Connection in the Import Data Excel dialog box. Customize the load
behavior by selecting Properties from the Home tab of Power Query Editor in
Power BI Desktop.

Power Pivot All Power Pivot external data connections convert to queries in Power BI Desktop.
external data
connections

Linked tables If there's a worksheet table in Excel linked to the data model, or linked to a query
or current (by using From Table or the Excel.CurrentWorkbook() function in M), you'll see the
workbook following options:
tables
Import the table to the Power BI Desktop file. This table is a one-time snapshot
of the data, after which the data is read-only in the table in Power BI Desktop.
There's a size limitation of 1 million characters (total, combining all column
headers and cells) for tables created using this option.

Keep a connection to the original workbook. Alternatively, you can keep a

connection to the original Excel Workbook. Power BI Desktop retrieves the latest
content in this table with each refresh, just like any other query you create
against an Excel workbook in Power BI Desktop.

Data model These data model objects convert to the equivalent objects in Power BI Desktop.
calculated Note there are certain data categories that aren't available in Power BI Desktop,
columns, such as Image. In these cases, the data category information resets for the
measures, columns in question.
KPIs, data
categories,
and
relationships
Are there any limitations to importing a
workbook?
There are a few limitations to importing a workbook into Power BI Desktop:

External connections to SQL Server Analysis Services tabular models: In Excel

2013, it's possible to create a connection to SQL Server Analysis Services tabular
models without the need to import the data. This type of connection isn't currently
supported as part of importing Excel workbooks into Power BI Desktop. As a
workaround, you must recreate these external connections in Power BI Desktop.
Hierarchies: This type of data model object isn't currently supported in Power BI
Desktop. As such, hierarchies are skipped as part of importing an Excel workbook
into Power BI Desktop.
Binary data columns: This type of data model column isn't currently supported in
Power BI Desktop. Binary data columns are removed from the resulting table in
Power BI Desktop.
Named Ranges using From Table in Power Query, or using
Excel.CurrentWorkbook in M: Importing this named range data into Power BI
Desktop isn't currently supported, but it's a planned update. Currently, these
named ranges are loaded into Power BI Desktop as a connection to the external
Excel workbook.
PowerPivot to SSRS: PowerPivot external connections to SQL Server Reporting
Services (SSRS) aren't currently supported, because that data source isn't currently
available in Power BI Desktop.
Connect to an Oracle database with
Power BI Desktop
Article • 07/26/2023

To connect to an Oracle database or Oracle Autonomous Database with Power BI

Desktop, install Oracle Client for Microsoft Tools (OCMT) on the computer running
Power BI Desktop. The OCMT software you use depends on which version of Power BI
Desktop you've installed: 32-bit or 64-bit. It also depends on your version of Oracle
server.

Supported Oracle Database versions:

Oracle Database 12c (12.1.0.2) and later

Oracle Autonomous Database - all versions

Determining which version of Power BI

Desktop is installed
To determine which version of Power BI Desktop is installed, on the Help ribbon, select
About, then check the Version line. In the following image, a 64-bit version of Power BI
Desktop is installed:

Install the Oracle Client for Microsoft Tools

Oracle Client for Microsoft Tools installs and configures Oracle Data Provider for .NET
(ODP.​NET) to support 32-bit and 64-bit Microsoft tool connections with Oracle on-
premises and cloud databases, including Oracle Autonomous Database. It is a graphical
installer that automates the Oracle Database Client setup process. It supports
connecting with Power BI Desktop, Power BI service, Excel, SQL Server Analysis Services,
SQL Server Data Tools, SQL Server Integration Services, SQL Server Reporting Services,
and BizTalk Server.

OCMT is free software. It can be downloaded from the Oracle Client for Microsoft Tools
page and is available for 32-bit or 64-bit Power BI Desktop.

Power BI Desktop uses unmanaged ODP.​NET to connect to Oracle database or Oracle

Autonomous Database.

Here are step by step instructions to use OCMT and setup Oracle database connectivity
to Power BI Desktop .

Connect to an Oracle database with on-

premises data gateway
Some Power BI Desktop app deployments use on-premises data gateway to connect to
Oracle database. To connect to an Oracle database with the on-premises data gateway,
use 64-bit OCMT on the computer running the gateway since the gateway is a 64-bit
app. For more information, go to Manage your data source - Oracle.

Connect to an Oracle Database

For information about connecting to an Oracle database or an Oracle Autonomous
database from either Power BI Desktop or Power BI service, go to the Power Query
article on Oracle databases.

Next steps
DirectQuery in Power BI
What is Power BI?
Data sources for the Power BI service
Oracle Client for Microsoft Tools

More questions? Ask the Power BI Community

Run Python scripts in Power BI Desktop
Article • 12/14/2022

You can run Python scripts directly in Power BI Desktop and import the resulting
datasets into a Power BI Desktop data model. From this model, you can create reports
and share them on the Power BI service.

Prerequisites
To run Python scripts in Power BI Desktop, you need to install Python on your local
machine. You can download Python from the Python website . The current
Python scripting release supports Unicode characters and spaces in the installation
path.

The Power BI Python integration requires installation of the following two Python
packages. In a console or shell, use the pip command-line tool to install the
packages. The pip tool is packaged with recent Python versions.

Pandas is a software library for data manipulation and analysis. Pandas offers
data structures and operations for manipulating numerical tables and time
series. To import into Power BI, Python data must be in a pandas data frame .
A data frame is a two-dimensional data structure, such as a table with rows and
columns.

Matplotlib is a plotting library for Python and its numerical mathematics

extension NumPy . Matplotlib provides an object-oriented API for embedding
plots into general-purpose graphical user interface (GUI) applications for
Python, such as Tkinter, wxPython, Qt, or GTK+.

Console

pip install pandas

pip install matplotlib

Enable Python scripting

To enable Python scripting in Power BI:

1. In Power BI Desktop, select File > Options and settings > Options > Python
scripting. The Python script options page appears.
2. If necessary, supply or edit your local Python installation path under Detected
Python home directories. In the preceding image, the Python's installation local
path is C:\Python. If you have more than one local Python installation, make sure to
select the one that you want to use.

3. Select OK.

） Important

Power BI runs scripts directly by using the python.exe executable from the directory
you provide in Settings. Python distributions that require an extra step to prepare
the environment, such as Conda, might fail to run. To avoid these issues, use the
official Python distribution from https://www.python.org . Another possible
solution is to start Power BI Desktop from your custom Python environment
prompt.
Create a Python script
Create a script in your local Python development environment and make sure it runs
successfully. To prepare and run a Python script in Power BI Desktop, there are a few
limitations:

Only pandas data frames import, so make sure the data you want to import to
Power BI is represented in a data frame.
Any Python script that runs longer than 30 minutes times out.
Interactive calls in the Python script, such as waiting for user input, halt the script's
execution.
If you set a working directory within the Python script, you must define a full path
to the working directory rather than a relative path.
Nested tables aren't supported.

Here's a simple example Python script that imports pandas and uses a data frame:

Python

import pandas as pd
data = [['Alex',10],['Bob',12],['Clarke',13]]
df = pd.DataFrame(data,columns=['Name','Age'],dtype=float)
print (df)

When run, this script returns:

Output

Name Age
0 Alex 10.0
1 Bob 12.0
2 Clarke 13.0

Run the script and import data

To run your Python script:

1. In the Home group of the Power BI Desktop ribbon, select Get data.

2. In the Get Data dialog box, select Other > Python script, and then select Connect.
Power BI uses your latest installed Python version as the Python engine.
3. On the Python script screen, paste your Python script into the Script field, and
select OK.
 4. If the script runs successfully, the Navigator window appears, and you can load the
data. Select the df table, and then select Load.

Power BI imports the data, and you can use it to create visualizations and reports. To
refresh the data, select Refresh in the Home group of the Power BI Desktop ribbon.
When you refresh, Power BI runs the Python script again.

） Important

If Python isn't installed or identified, a warning appears. You might also get a
warning if you have multiple local machine installations.
Next steps
For more information about Python in Power BI, see:

Create Python visuals in Power BI Desktop

Use an external Python IDE with Power BI
Use Python in Power Query Editor
Use Python in Power Query Editor
Article • 02/13/2023

You can use Python, a programming language widely used by statisticians, data
scientists, and data analysts, in the Power BI Desktop Power Query Editor. This
integration of Python into Power Query Editor lets you perform data cleansing using
Python, and perform advanced data shaping and analytics in datasets, including
completion of missing data, predictions, and clustering, just to name a few. Python is a
powerful language, and can be used in Power Query Editor to prepare your data model
and create reports.

Prerequisites
You'll need to install Python and pandas before you begin.

Install Python - To use Python in Power BI Desktop's Power Query Editor, you need
to install Python on your local machine. You can download and install Python for
free from many locations, including the Official Python download page , and the
Anaconda .

Install pandas - To use Python with the Power Query Editor, you'll also need to
install pandas . Pandas is used to move data between Power BI and the Python
environment.

Use Python with Power Query Editor

To show how to use Python in Power Query Editor, take this example from a stock
market dataset, based on a CSV file that you can download from here and follow
along. The steps for this example are the following procedure:

1. First, load your data into Power BI Desktop. In this example, load the
EuStockMarkets_NA.csv file and select Get data > Text/CSV from the Home ribbon
in Power BI Desktop.
2. Select the file and select Open, and the CSV is displayed in the CSV file dialog.

3. Once the data is loaded, you see it in the Fields pane in Power BI Desktop.
4. Open Power Query Editor by selecting Transform data from the Home tab in
Power BI Desktop.

5. In the Transform tab, select Run Python Script and the Run Python Script editor
appears as shown in the next step. Rows 15 and 20 suffer from missing data, as do
other rows you can't see in the following image. The following steps show how
Python completes those rows for you.
6. For this example, enter the following script code:

Python

import pandas as pd
completedData = dataset.fillna(method='backfill', inplace=False)
dataset["completedValues"] = completedData["SMI missing values"]

７ Note

You need to have the pandas library installed in your Python environment for
the previous script code to work properly. To install pandas, run the following
command in your Python installation: pip install pandas

When put into the Run Python Script dialog, the code looks like the following
example:

7. After you select OK, Power Query Editor displays a warning about data privacy.
 8. For the Python scripts to work properly in the Power BI service, all data sources
need to be set to public. For more information about privacy settings and their
implications, see Privacy Levels.

Notice a new column in the Fields pane called completedValues. Notice there are a
few missing data elements, such as on row 15 and 18. Take a look at how Python
handles that in the next section.

With just three lines of Python script, Power Query Editor filled in the missing values
with a predictive model.

Create visuals from Python script data

Now we can create a visual to see how the Python script code using the pandas library
completed the missing values, as shown in the following image:
Once that visual is complete, and any other visuals you might want to create using
Power BI Desktop, you can save the Power BI Desktop file. Power BI Desktop files save
with the .pbix file name extension. Then use the data model, including the Python scripts
that are part of it, in the Power BI service.

７ Note

Want to see a completed .pbix file with these steps completed? You're in luck. You
can download the completed Power BI Desktop file used in these examples right
here .

Once you upload the .pbix file to the Power BI service, a couple more steps are
necessary to enable data to refresh in the service and to enable visuals to be updated in
the service. The data needs access to Python for visuals to be updated. The other steps
are the following steps:

Enable scheduled refresh for the dataset. To enable scheduled refresh for the
workbook that contains your dataset with Python scripts, see Configuring
scheduled refresh, which also includes information about Personal Gateway.
Install the Personal Gateway. You need a Personal Gateway installed on the
machine where the file is located, and where Python is installed. The Power BI
service must access that workbook and re-render any updated visuals. For more
information, see install and configure Personal Gateway.

Considerations and limitations

There are some limitations to queries that include Python scripts created in Power
Query Editor:
All Python data source settings must be set to Public, and all other steps in a query
created in Power Query Editor must also be public. To get to data source settings,
in Power BI Desktop select File > Options and settings > Data source settings.

From the Data Source Settings dialog, select the data sources and then select Edit
Permissions... and ensure that the Privacy Level is set to Public.
 To enable scheduled refresh of your Python visuals or dataset, you need to enable
Scheduled refresh and have a Personal Gateway installed on the computer that
houses the workbook and the Python installation. For more information on both,
see the previous section in this article, which provides links to learn more about
each.

Nested tables, which are table of tables, are currently not supported.

There are all sorts of things you can do with Python and custom queries, so explore and
shape your data just the way you want it to appear.
Use an external Python IDE with Power
BI
Article • 01/13/2023

With Power BI Desktop, you can use your external Python Integrated Development
Environment (IDE) to create and refine Python scripts, then use those scripts in Power BI.

Enable an external Python IDE

You can launch your external Python IDE from Power BI Desktop and have your data
automatically imported and displayed in the Python IDE. From there, you can modify the
script in that external Python IDE, then paste it back into Power BI Desktop to create
Power BI visuals and reports.
You can specify which Python IDE to use, and have it launch automatically from within
Power BI Desktop.

Requirements
To use this feature, you need to install a Python IDE on your local computer. Power BI
Desktop doesn't include, deploy, or install the Python engine, so you must separately
install Python on your local computer. You can choose which Python IDE to use, with the
following options:

You can install your favorite Python IDE, many of which are available for free, such
as the Visual Studio Code download page .

Power BI Desktop also supports Visual Studio.

You can also install a different Python IDE and have Power BI Desktop launch that
Python IDE by doing one of the following:
You can associate .PY files with the external IDE you want Power BI Desktop to
launch.
You can specify the .exe that Power BI Desktop launches by selecting Other from
the Python script options section of the Options dialog. You can bring up the
Options dialog by going to File > Options and settings > Options.
If you have multiple Python IDEs installed, you can specify which is launched by
selecting it from the Detected Python IDEs drop-down in the Options dialog.

By default, Power BI Desktop launches Visual Studio Code as the external Python IDE if
it's installed on your local computer. If Visual Studio Code isn't installed and you have
Visual Studio that is launched instead. If neither of those Python IDEs is installed, the
application associated with .PY files is launched.

And if no .PY file association exists, it's possible to specify a path to a custom IDE in the
Set a Python home directory section of the Options dialog. You can also launch a
different Python IDE by selecting the Settings gear icon beside the Launch Python IDE
arrow icon, in Power BI Desktop.

Launch a Python IDE from Power BI Desktop

To launch a Python IDE from Power BI Desktop, take the following steps:

1. Load data into Power BI Desktop.

2. Add a Python visualization to your canvas. If you haven't enabled script visuals yet,
you're prompted to do so.

3. After script visuals are enabled, a blank Python visual appears that's ready to
display the results of your script. The Python script editor pane also appears.

4. Now you can select the fields you want to use in your Python script. When you
select a field, the Python script editor field automatically creates script code based
on the field or fields you select. You can either create or paste your Python script
directly in the Python script editor pane, or you can leave it empty.
 

７ Note

The default aggregation type for Python visuals is do not summarize.

5. You can now launch your Python IDE directly from Power BI Desktop. Select the
Launch Python IDE button, found on the right side of the Python script editor title
bar, as shown in this screenshot.

6. Your specified Python IDE is launched by Power BI Desktop, as shown in the

following image. In this image, Visual Studio Code is the default Python IDE.
 

７ Note

Power BI Desktop adds the first three lines of the script so it can import your
data from Power BI Desktop once you run the script.

7. Any script you created in the Python script editor pane of Power BI Desktop
appears, starting in line 4, in your Python IDE. At this point, you can create your
Python script in the Python IDE. Once your Python script is complete in your
Python IDE, you need to copy and paste it back into the Python script editor pane
in Power BI Desktop, excluding the first three lines of the script that Power BI
Desktop automatically generated. Don't copy the first three lines of script back into
Power BI Desktop, those lines were only used to import your data to your Python
IDE from Power BI Desktop.

Known limitations
Launching a Python IDE directly from Power BI Desktop has a few limitations:

Automatically exporting your script from your Python IDE into Power BI Desktop
isn't supported.
Next steps
Take a look at the following additional information about Python in Power BI.

Running Python Scripts in Power BI Desktop

Create Power BI visuals using Python
Create Power BI visuals with Python
Article • 01/18/2023

This tutorial helps you get started creating visuals with Python data in Power BI Desktop.
You use a few of the many available options and capabilities for creating visual reports
by using Python, pandas, and the Matplotlib library.

Prerequisites
Work through Run Python scripts in Power BI Desktop to:

Install Python on your local machine.

Enable Python scripting in Power BI Desktop.

Install the pandas and Matplotlib Python libraries.

Import the following Python script into Power BI Desktop:

Python

import pandas as pd
df = pd.DataFrame({
'Fname':['Harry','Sally','Paul','Abe','June','Mike','Tom'],
'Age':[21,34,42,18,24,80,22],
'Weight': [180, 130, 200, 140, 176, 142, 210],
'Gender':['M','F','M','M','F','M','M'],
'State':
['Washington','Oregon','California','Washington','Nevada','Texas','Neva
da'],
'Children':[4,1,2,3,0,2,0],
'Pets':[3,2,2,5,0,1,5]
})
print (df)

Create a Python visual in Power BI Desktop

1. After you import the Python script, select the Python visual icon in the Power BI
Desktop Visualizations pane.
2. In the Enable script visuals dialog box that appears, select Enable.

A placeholder Python visual image appears on the report canvas, and the Python
script editor appears along the bottom of the center pane.

3. Drag the Age, Children, Fname, Gender, Pets, State, and Weight fields to the
Values section where it says Add data fields here.
 Based on your selections, the Python script editor generates the following binding
code.

The editor creates a dataset dataframe with the fields you add.
The default aggregation is Don't summarize.
Similar to table visuals, fields are grouped and duplicate rows appear only
once.

4. With the dataframe automatically generated by the fields you selected, you can
write a Python script that results in plotting to the Python default device. When the
script is complete, select the Run icon from the Python script editor title bar to run
the script and generate the visual.

Tips
 Your Python script can use only fields that are added to the Values section. You can
add or remove fields while you work on your Python script. Power BI Desktop
automatically detects field changes. As you select or remove fields from the Values
section, supporting code in the Python script editor is automatically generated or
removed.

In some cases, you might not want automatic grouping to occur, or you might
want all rows to appear, including duplicates. In those cases, you can add an index
field to your dataset that causes all rows to be considered unique and prevents
grouping.

You can access columns in the dataset by using their names. For example, you can
code dataset["Age"] in your Python script to access the age field.

Power BI Desktop replots the visual when you select Run from the Python script
editor title bar, or whenever a data change occurs due to data refresh, filtering, or
highlighting.

When you run a Python script that results in an error, the Python visual isn't
plotted, and an error message appears on the canvas. For error details, select See
details in the message.

To get a larger view of the visualizations, you can minimize the Python script
editor.

Create a scatter plot

Create a scatter plot to see if there's a correlation between age and weight.

1. In the Python script editor, under Paste or type your script code here, enter this
code:

Python

import matplotlib.pyplot as plt

dataset.plot(kind='scatter', x='Age', y='Weight', color='red')
plt.show()

Your Python script editor pane should now look like the following image:
 The code imports the Matplotlib library, which plots and creates the visual.

2. Select the Run script button to generate the following scatter plot in the Python
visual.

Create a line plot with multiple columns

Create a line plot for each person that shows their number of children and pets.

1. Under Paste or type your script code here, remove or comment out the previous
code, and enter the following Python code:

Python

import matplotlib.pyplot as plt

ax = plt.gca()
dataset.plot(kind='line',x='Fname',y='Children',ax=ax)
dataset.plot(kind='line',x='Fname',y='Pets', color='red', ax=ax)
plt.show()

2. Select the Run button to generate the following line plot with multiple columns:
Create a bar plot
Create a bar plot for each person's age.

1. Under Paste or type your script code here, remove or comment out the previous
code, and enter the following Python code:

Python

import matplotlib.pyplot as plt

dataset.plot(kind='bar',x='Fname',y='Age')
plt.show()

2. Select the Run button to generate the following bar plot:

Limitations
Python visuals in Power BI Desktop have the following limitations:

The data the Python visual uses for plotting is limited to 150,000 rows. If more than
150,000 rows are selected, only the top 150,000 rows are used, and a message
appears on the image. The input data also has a limit of 250 MB.

If the input dataset of a Python visual has a column that contains a string value
longer than 32,766 characters, that value is truncated.

All Python visuals display at 72 DPI resolution.

If a Python visual calculation exceeds five minutes, the execution times out, which
results in an error.

As with other Power BI Desktop visuals, if you select data fields from different
tables with no defined relationship between them, an error occurs.

Python visuals refresh upon data updates, filtering, and highlighting. The image
itself isn't interactive.

Python visuals respond to highlighting elements in other visuals, but you can't
select elements in the Python visual to cross-filter other elements.

Only plots to the Python default display device display correctly on the canvas.
Avoid explicitly using a different Python display device.
 Python visuals don't support renaming input columns. Columns are referred to by
their original names during script execution.

Security
Python visuals use Python scripts, which could contain code that has security or privacy
risks. When you attempt to view or interact with a Python visual for the first time, you
get a security warning. Enable Python visuals only if you trust the author and source, or
after you review and understand the Python script.

Licensing
Python visuals require a Power BI Pro or Premium Per User (PPU) license to render in
reports, refresh, filter, and cross-filter. Users of free Power BI can consume only tiles that
are shared with them in Premium workspaces.

The following table describes Python visuals capabilities based on licensing.

Author Python Create Power BI service View Python

visuals in Power BI reports with Python visuals in reports
Desktop visuals

Guest (Power BI Supported Not supported Supported in

embedded) Premium/Azure
capacity only

Unmanaged tenant Supported Not supported Not supported

(domain not
verified)

Managed tenant Supported Not supported Supported in

with free license Premium capacity
only

Managed tenant Supported Supported Supported

with Pro or PPU
license

For more information about Power BI Pro licenses and how they differ from free licenses,
see Purchase and assign Power BI Pro user licenses.

Next steps
This tutorial barely scratches the surface of the options and capabilities for creating
visual reports using by Python, pandas, and the Matplotlib library. For more information,
see the following resources:

Documentation at the Matplotlib website.

Matplotlib Tutorial : A Basic Guide to Use Matplotlib with Python
Matplotlib Tutorial – Python Matplotlib Library with Examples
Pandas API Reference
Python visualizations in Power BI service
Using Python Visuals in Power BI
Comprehensive Python Scripting Tutorial

For more information about Python in Power BI, see:

Run Python Scripts in Power BI Desktop

Use an external Python IDE with Power BI
Learn which Python packages are
supported in Power BI
Article • 01/23/2024

You can use the powerful Python programming language to create visuals in Power BI.
Many Python packages are supported in Power BI and more are being supported all the
time.

The following sections provide an alphabetical table of which Python packages are
supported in Power BI.

Request support for a new Python package

Supported Python packages for Power BI are found in the following section. If you
would like to request support of a Python package not found in that list, submit your
request to Power BI Ideas .

Requirements and limitations of Python

packages
There are a handful of requirements and limitations for Python packages:

Current Python runtime: Python 3.7.7.

Power BI, for the most part, supports Python packages with free and open-source
software licenses such as GPL-2, GPL-3, MIT+, and so on.
Power BI supports packages published in PyPI. The service doesn't support private
or custom Python packages. Users are encouraged to make their private packages
available on PyPI prior to requesting the package be available in Power BI.
For Python visuals in Power BI Desktop, you can install any package, including
custom Python packages.
For security and privacy reasons, Python packages that provide client-server
queries over the web in the service, aren't supported. Networking is blocked for
such attempts.
The approval process for including a new Python package has a tree of
dependencies. Some dependencies required to be installed in the service can't be
supported.
Python packages that are supported in Power
BI
The following table shows which packages are supported in Power BI.

ﾉ Expand table

Package Version Link

cycler 0.11.0 https://pypi.org/project/cycler

joblib 1.1.0 https://pypi.org/project/joblib

kiwisolver 1.4.4 https://pypi.org/project/kiwisolver

matplotlib 3.2.2 https://pypi.org/project/matplotlib

numpy 1.21.6 https://pypi.org/project/numpy

packaging 21.3 https://pypi.org/project/packaging

pandas 1.3.5 https://pypi.org/project/pandas

patsy 0.5.2 https://pypi.org/project/patsy

pip 22.1.2 https://pypi.org/project/pip

pyparsing 3.0.9 https://pypi.org/project/pyparsing

python-dateutil 2.8.2 https://pypi.org/project/python-dateutil

pytz 2022.1 https://pypi.org/project/pytz

scikit-learn 1.0.2 https://pypi.org/project/scikit-learn

scipy 1.7.3 https://pypi.org/project/scipy

seaborn 0.11.2 https://pypi.org/project/seaborn

setuptools 63.2.0 https://pypi.org/project/setuptools

six 1.16.0 https://pypi.org/project/six

statsmodels 0.13.2 https://pypi.org/project/statsmodels

threadpoolctl 3.1.0 https://pypi.org/project/threadpoolctl

typing-extensions 4.3.0 https://pypi.org/project/typing-extensions

xgboost 1.6.1 https://pypi.org/project/xgboost

Related content
For more information about Python in Power BI, take a look at the following articles:

Create Power BI visuals using Python

Running Python scripts in Power BI Desktop
Using Python in Query Editor
Run R scripts in Power BI Desktop
Article • 11/10/2023

You can run R scripts directly in Power BI Desktop and import the resulting semantic
models into a Power BI Desktop data model.

Install R
To run R scripts in Power BI Desktop, you need to install R on your local machine. You
can download and install R for free from many locations, including the CRAN
Repository . The current release supports Unicode characters and spaces (empty
characters) in the installation path.

Run R scripts
Using just a few steps in Power BI Desktop, you can run R scripts and create a data
model. With the data model, you can create reports and share them on the Power BI
service. R scripting in Power BI Desktop now supports number formats that contain
decimals (.) and commas (,).

Prepare an R script
To run an R script in Power BI Desktop, create the script in your local R development
environment, and make sure it runs successfully.

To run the script in Power BI Desktop, make sure the script runs successfully in a new
and unmodified workspace. This prerequisite means that all packages and dependencies
must be explicitly loaded and run. You can use source() to run dependent scripts.

When you prepare and run an R script in Power BI Desktop, there are a few limitations:

Because only data frames are imported, remember to represent the data you want
to import to Power BI in a data frame.
Columns typed as Complex and Vector aren't imported, and they're replaced with
error values in the created table.
Values of N/A are translated to NULL values in Power BI Desktop.
If an R script runs longer than 30 minutes, it times out.
Interactive calls in the R script, such as waiting for user input, halt the script's
execution.
 When setting the working directory within the R script, you must define a full path
to the working directory, rather than a relative path.
R scripts cannot run in the Power BI service.

Run your R script and import data

Now you can run your R script to import data into Power BI Desktop:

1. In Power BI Desktop, select Get data, choose Other > R script, and then select
Connect:

2. If R is installed on your local machine, just copy your script into the script window
and select OK. The latest installed version is displayed as your R engine.
 3. Select OK to run the R Script. When the script runs successfully, you can then
choose the resulting data frames to add to the Power BI model.

You can control which R installation to use to run your script. To specify your R
installation settings, choose File > Options and settings > Options, then select R
scripting. Under R script options, the Detected R home directories dropdown list shows
your current R installation choices. If the R installation you want isn't listed, pick Other,
and then browse to or enter your preferred R installation folder in Set an R home
directory.
Refresh
You can refresh an R script in Power BI Desktop. When you refresh an R script, Power BI
Desktop runs the R script again in the Power BI Desktop environment.

Next steps
Take a look at the following additional information about R in Power BI.

Create Power BI visuals using R

Use an external R IDE with Power BI
Use R in Power Query Editor
Article • 11/20/2023

The R language is a powerful programming language that many statisticians, data

scientists, and data analysts use. You can use R in Power BI Desktop's Power Query
Editor to:

Prepare data models.

Create reports.
Do data cleansing, advanced data shaping, and semantic model analytics, which
include missing data completion, predictions, clustering, and more.

Install R
You can download R for free from the CRAN Repository .

Install mice
As a prerequisite, you must install the mice library in your R environment. Without
mice, the sample script code doesn't work properly. The mice package implements a
method to deal with missing data.

To install the mice library:

1. Launch the R.exe program, for example, C:\Program Files\Microsoft\R Open\R-

3.5.3\bin\R.exe .

2. Run the install command from the R prompt:

install.packages('mice')

Use an R script in Power Query Editor

To demonstrate using R in Power Query Editor, this example uses a stock market
semantic model contained in a .csv file.

1. Download the EuStockMarkets_NA.csv file . Remember where you save it.

2. Load the file into Power BI Desktop. From the Home tab, select Get data >
Text/CSV.

3. Select the EuStockMarkets_NA.csv file, and then choose Open. The CSV data is
displayed in the Text/CSV file dialog.

4. Select Load to load the data from the file. After Power BI Desktop has loaded the
data, the new table appears in the Fields pane.
5. To open Power Query Editor, from the Home ribbon select Transform data.

6. From the Transform tab, select Run R script. The Run R script editor appears. Rows
15 and 20 have missing data, as do other rows you can't see in the image. The
following steps show how R completes those rows for you.

7. For this example, enter the following script code in the Script box of the Run R
script window.

R
 library(mice)
tempData <- mice(dataset,m=1,maxit=50,meth='pmm',seed=100)
completedData <- complete(tempData,1)
output <- dataset
output$completedValues <- completedData$"SMI missing values"

７ Note

You might need to overwrite a variable named output to properly create the
new semantic model with the filters applied.

8. Select OK. Power Query Editor displays a warning about data privacy.

9. Inside the warning message, select Continue. In the Privacy levels dialog that
appears, set all data sources to Public for the R scripts to work properly in the
Power BI service.
 For more information about privacy settings and their implications, see Power BI
Desktop privacy levels.

10. Select Save to run the script.

When you run the script, you see the following result:

When you select Table next to Output in the table that appears, the table is
presented, as shown in the following image.

Notice the new column in the Fields pane called completedValues. The SMI
missing values column has a few missing data elements. Take a look at how R
handles that in the next section.

With just five lines of R script, Power Query Editor filled in the missing values with a
predictive model.

Create visuals from R script data

We can now create a visual to see how the R script code with the mice library completes
the missing values.
 

You can save all completed visuals in one Power BI Desktop .pbix file and use the data
model and its R scripts in the Power BI service.

７ Note

You can download a .pbix file with all these steps completed.

After you've uploaded the .pbix file to the Power BI service, you need to take other steps
to enable service data refresh and updated visuals:

Enable scheduled refresh for the semantic model: To enable scheduled refresh for
the workbook containing your semantic model with R scripts, see Configuring
scheduled refresh. This article also includes information about on-premises data
gateways.

Install a gateway: You need an on-premises data gateway (personal mode)

installed on the machine where the file and R are located. The Power BI service
accesses that workbook and re-renders any updated visuals. For more information,
see use personal gateways in Power BI.

Considerations and limitations

There are some limitations to queries that include R scripts created in Power Query
Editor:
All R data source settings must be set to Public. All other steps in a Power Query
Editor query must also be public.

To get to the data source settings, in Power BI Desktop, select File > Options and
settings > Data source settings.

In the Data source settings dialog, select one or more data sources, and then
select Edit Permissions. Set the Privacy Level to Public.
 

To schedule refresh of your R visuals or semantic model, enable scheduled refresh

and install an on-premises data gateway (personal mode) on the computer
containing the workbook and R. You can't use an enterprise gateway to refresh
semantic models containing R scripts in Power Query.

Next Steps
There are all sorts of things you can do with R and custom queries. Explore and shape
your data just the way you want it to appear.

Run R scripts in Power BI Desktop

Use an external R IDE with Power BI
Create visuals by using R packages in the Power BI service
Use an external R IDE with Power BI
Article • 12/07/2021

With Power BI Desktop, you can use your external R IDE (Integrated Development
Environment) to create and refine R scripts, then use those scripts in Power BI.

Enable an external R IDE

Launch your external R IDE from Power BI Desktop and have your data automatically
imported and displayed in the R IDE. From there, you can modify the script in that
external R IDE, then paste it back into Power BI Desktop to create Power BI visuals and
reports. Specify which R IDE you would like to use, and have it launch automatically from
within Power BI Desktop.
Requirements
To use this feature, you need to install an R IDE on your local computer. Power BI
Desktop does not include, deploy, or install the R engine, so you must separately install
R on your local computer. You can choose which R IDE to use, with the following
options:

You can install your favorite R IDE, many of which are available for free, such as the
Revolution Open download page , and the CRAN Repository .

Power BI Desktop also supports R Studio and Visual Studio 2015 with R Tools for
Visual Studio editors.

You can also install a different R IDE and have Power BI Desktop launch that R IDE
by doing one of the following:

You can associate .R files with the external IDE you want Power BI Desktop to
launch.

You can specify the .exe that Power BI Desktop should launch by selecting
Other from the R Script Options section of the Options dialog. You can bring up
the Options dialog by going to File > Options and settings > Options.
If you have multiple R IDEs installed, you can specify which will be launched by selecting
it from the Detected R IDEs drop-down in the Options dialog.

By default, Power BI Desktop will launch R Studio as the external R IDE if it's installed on
your local computer; if R Studio is not installed and you have Visual Studio 2015 with R
Tools for Visual Studio, that will be launched instead. If neither of those R IDEs is
installed, the application associated with .R files is launched.

And if no .R file association exists, it's possible to specify a path to a custom IDE in the
Browse to your preferred R IDE section of the Options dialog. You can also launch a
different R IDE by selecting the Settings gear icon beside the Edit script in external IDE
arrow icon, in Power BI Desktop.

Launch an R IDE from Power BI Desktop

To launch an R IDE from Power BI Desktop, take the following steps:

1. Load data into Power BI Desktop.

2. When script visuals are enabled, you can select an R visual from the Visualizations
pane, which creates a blank R visual that's ready to display the results of your
script. The R script editor pane also appears.

3. Select some fields from the Fields pane that you want to work with. If you haven't
enabled script visuals yet, you'll be prompted to do so.

4. Now you can select the fields you want to use in your R script. When you select a
field, the R script editor field automatically creates script code based on the field
or fields you select. You can either create (or paste) your R script directly in the R
script editor pane, or you can leave it empty.
 ７ Note

The default aggregation type for R visuals is do not summarize.

5. You can now launch your R IDE directly from Power BI Desktop. Select the Edit
script in external IDE button, found on the right side of the R script editor title
bar, as shown below.

6. Your specified R IDE is launched by Power BI Desktop, as shown in the following

image (in this image, RStudio is the default R IDE).
 ７ Note

Power BI Desktop adds the first three lines of the script so it can import your
data from Power BI Desktop once you run the script.

7. Any script you created in the R script editor pane of Power BI Desktop appears
starting in line 4 in your R IDE. At this point, you can create your R script in the R
IDE. Once your R script is complete in your R IDE, you need to copy and paste it
back into the R script editor pane in Power BI Desktop, excluding the first three
lines of the script that Power BI Desktop automatically generated. Do not copy the
first three lines of script back into Power BI Desktop, those lines were only used to
import your data to your R IDE from Power BI Desktop.

Known limitations
Launching an R IDE directly from Power BI Desktop has a few limitations:

Automatically exporting your script from your R IDE into Power BI Desktop is not
supported.
R Client editor (RGui.exe) is not supported, because the editor itself does not
support opening files.

Next steps
Take a look at the following additional information about R in Power BI.
Running R Scripts in Power BI Desktop
Create Power BI visuals using R
Create visuals by using R packages in
the Power BI service
Article • 11/10/2023

You can use the powerful R programming language to create visuals in the Power BI
service. Many R packages are supported in the Power BI service and more are being
supported all the time. Some packages aren't supported.

The following sections provide an alphabetical table of which R packages are supported
in Power BI, and which aren't. For more information about R in Power BI, see the R
visuals article.

Request support for a new R package

Supported R packages for the Power BI service are found in the following section. If you
would like to request support of an R package not found in that list, submit your request
to Power BI Ideas .

Requirements and limitations of R packages

There are a handful of requirements and limitations for R packages:

Current R runtime: Microsoft R 3.4.4

The Power BI service usually supports R packages with free and open-source
software licenses such as GPL-2, GPL-3, MIT+, and so on.

The Power BI service supports packages published in the Comprehensive R Archive

Network (CRAN). The service doesn't support private or custom R packages. Users
are encouraged to make their private packages available on CRAN prior to
requesting the package be available in the Power BI service.

The Power BI Desktop has two variations for R packages:

For R visuals, you can install any package, including custom R packages
For Custom R visuals, only public CRAN packages are supported for auto-
installation of the packages

For security and privacy reasons, R packages that provide client-server queries over
the web, such as RgoogleMaps, in the service, aren't supported. Networking is
 blocked for such attempts. See the following section for a list of supported and
unsupported R packages.

The approval process for including a new R package has a tree of dependencies.
Some dependencies required to be installed in the service can't be supported.

R packages that are supported in Power BI

The following table shows which packages are supported in the Power BI service.

Package Version Link

abc 2.1 https://cran.r-project.org/web/packages/abc/index.html

abc.data 1.0 https://cran.r-project.org/web/packages/abc.data/index.html

abind 1.4-5 https://cran.r-project.org/web/packages/abind/index.html

acepack 1.4.1 https://cran.r-project.org/web/packages/acepack/index.html

actuar 2.3-1 https://cran.r-project.org/web/packages/actuar/index.html

ade4 1.7-10 https://cran.r-project.org/web/packages/ade4/index.html

adegenet 2.1.2 https://cran.r-project.org/web/packages/adegenet/index.html

AdMit 2.1.3 https://cran.r-project.org/web/packages/AdMit/index.html

AER 1.2-5 https://cran.r-project.org/web/packages/AER/index.html

agricolae 1.3-1 https://cran.r-project.org/web/packages/agricolae/index.html

AlgDesign 1.1-7.3 https://cran.r-project.org/web/packages/AlgDesign/index.html

alluvial 0.1-2 https://cran.r-project.org/web/packages/alluvial/index.html

andrews 1.0 https://cran.r-project.org/web/packages/andrews/index.html

anomalize 0.1.1 https://cran.r-project.org/web/packages/anomalize/index.html

anytime 0.3.3 https://cran.r-project.org/web/packages/anytime/index.html

aod 1.3 https://cran.r-project.org/web/packages/aod/index.html

apcluster 1.4.5 https://cran.r-project.org/web/packages/apcluster/index.html

ape 5.0 https://cran.r-project.org/web/packages/ape/index.html

aplpack 1.3.0 https://cran.r-project.org/web/packages/aplpack/index.html

Package Version Link

approximator 1.2-6 https://cran.r-

project.org/web/packages/approximator/index.html

arm 1.9-3 https://cran.r-project.org/web/packages/arm/index.html

arules 1.6-0 https://cran.r-project.org/web/packages/arules/index.html

arulesViz 1.3-0 https://cran.r-project.org/web/packages/arulesViz/index.html

ash 1.0-15 https://cran.r-project.org/web/packages/ash/index.html

assertthat 0.2.0 https://cran.r-project.org/web/packages/assertthat/index.html

autocogs 0.1.2 https://cran.r-project.org/web/packages/autocogs/index.html

automap 1.0-14 https://cran.r-project.org/web/packages/automap/index.html

aweek 1.0.1 https://cran.r-project.org/web/packages/aweek/index.html

AzureML 0.2.14 https://cran.r-project.org/web/packages/AzureML/index.html

BaBooN 0.2-0 https://cran.r-project.org/web/packages/BaBooN/index.html

BACCO 2.0-9 https://cran.r-project.org/web/packages/BACCO/index.html

backports 1.1.2 https://cran.r-project.org/web/packages/backports/index.html

BaM 1.0.1 https://cran.r-project.org/web/packages/BaM/index.html

BAS 1.4.9 https://cran.r-project.org/web/packages/BAS/index.html

base 3.4.4 NA

base2grob 0.0.2 https://cran.r-project.org/web/packages/base2grob/index.html

base64 2.0 https://cran.r-project.org/web/packages/base64/index.html

base64enc 0.1-3 https://cran.r-project.org/web/packages/base64enc/index.html

BayesDA 2012.04-1 https://cran.r-project.org/web/packages/BayesDA/index.html

BayesFactor 0.9.12-2 https://cran.r-

project.org/web/packages/BayesFactor/index.html

bayesGARCH 2.1.3 https://cran.r-

project.org/web/packages/bayesGARCH/index.html

bayesm 3.1-0.1 https://cran.r-project.org/web/packages/bayesm/index.html

bayesmix 0.7-4 https://cran.r-project.org/web/packages/bayesmix/index.html

Package Version Link

bayesplot 1.5.0 https://cran.r-project.org/web/packages/bayesplot/index.html

bayesQR 2.3 https://cran.r-project.org/web/packages/bayesQR/index.html

bayesSurv 3.2 https://cran.r-project.org/web/packages/bayesSurv/index.html

Bayesthresh 2.0.1 https://cran.r-

project.org/web/packages/Bayesthresh/index.html

BayesTree 0.3-1.4 https://cran.r-project.org/web/packages/BayesTree/index.html

BayesValidate 0.0 https://cran.r-

project.org/web/packages/BayesValidate/index.html

BayesX 0.2-9 https://cran.r-project.org/web/packages/BayesX/index.html

BayHaz 0.1-3 https://cran.r-project.org/web/packages/BayHaz/index.html

bbemkr 2.0 https://cran.r-project.org/web/packages/bbemkr/index.html

BCBCSF 1.0-1 https://cran.r-project.org/web/packages/BCBCSF/index.html

BCE 2.1 https://cran.r-project.org/web/packages/BCE/index.html

bclust 1.5 https://cran.r-project.org/web/packages/bclust/index.html

bcp 4.0.0 https://cran.r-project.org/web/packages/bcp/index.html

BDgraph 2.45 https://cran.r-project.org/web/packages/BDgraph/index.html

beanplot 1.2 https://cran.r-project.org/web/packages/beanplot/index.html

beeswarm 0.2.3 https://cran.r-project.org/web/packages/beeswarm/index.html

benford.analysis 0.1.4.1 https://cran.r-

project.org/web/packages/benford.analysis/index.html

BenfordTests 1.2.0 https://cran.r-

project.org/web/packages/BenfordTests/index.html

bfp 0.0-38 https://cran.r-project.org/web/packages/bfp/index.html

BH 1.66.0-1 https://cran.r-project.org/web/packages/BH/index.html

biglm 0.9-1 https://cran.r-project.org/web/packages/biglm/index.html

bindr 0.1.1 https://cran.r-project.org/web/packages/bindr/index.html

bindrcpp 0.2.2 https://cran.r-project.org/web/packages/bindrcpp/index.html

binom 1.1-1 https://cran.r-project.org/web/packages/binom/index.html

Package Version Link

bisoreg 1.4 https://cran.r-project.org/web/packages/bisoreg/index.html

bit 1.1-12 https://cran.r-project.org/web/packages/bit/index.html

bit64 0.9-7 https://cran.r-project.org/web/packages/bit64/index.html

bitops 1.0-6 https://cran.r-project.org/web/packages/bitops/index.html

bizdays 1.0.6 https://cran.r-project.org/web/packages/bizdays/index.html

blandr 0.5.1 https://cran.r-project.org/web/packages/blandr/index.html

blme 1.0-4 https://cran.r-project.org/web/packages/blme/index.html

blob 1.1.1 https://cran.r-project.org/web/packages/blob/index.html

BLR 1.4 https://cran.r-project.org/web/packages/BLR/index.html

BMA 3.18.8 https://cran.r-project.org/web/packages/BMA/index.html

Bmix 0.6 https://cran.r-project.org/web/packages/Bmix/index.html

bmp 0.3 https://cran.r-project.org/web/packages/bmp/index.html

BMS 0.3.4 https://cran.r-project.org/web/packages/BMS/index.html

bnlearn 4.3 https://cran.r-project.org/web/packages/bnlearn/index.html

boa 1.1.8-2 https://cran.r-project.org/web/packages/boa/index.html

bomrang 0.1.4 https://cran.r-project.org/web/packages/bomrang/index.html

BoolNet 2.1.5 https://cran.r-project.org/web/packages/BoolNet/index.html

Boom 0.7 https://cran.r-project.org/web/packages/Boom/index.html

BoomSpikeSlab 0.9.0 https://cran.r-

project.org/web/packages/BoomSpikeSlab/index.html

boot 1.3-20 https://cran.r-project.org/web/packages/boot/index.html

bootstrap 2017.2 https://cran.r-project.org/web/packages/bootstrap/index.html

Boruta 5.3.0 https://cran.r-project.org/web/packages/Boruta/index.html

bqtl 1.0-32 https://cran.r-project.org/web/packages/bqtl/index.html

BradleyTerry2 1.0-8 https://cran.r-

project.org/web/packages/BradleyTerry2/index.html

brew 1.0-6 https://cran.r-project.org/web/packages/brew/index.html

Package Version Link

brglm 0.6.1 https://cran.r-project.org/web/packages/brglm/index.html

broom 0.4.4 https://cran.r-project.org/web/packages/broom/index.html

bspec 1.5 https://cran.r-project.org/web/packages/bspec/index.html

bspmma 0.1-1 https://cran.r-project.org/web/packages/bspmma/index.html

bsts 0.7.1 https://cran.r-project.org/web/packages/bsts/index.html

bupaR 0.4.4 https://cran.r-project.org/web/packages/bupaR/index.html

BVS 4.12.1 https://cran.r-project.org/web/packages/BVS/index.html

C50 0.1.1 https://cran.r-project.org/web/packages/C50/index.html

Cairo 1.5-9 https://cran.r-project.org/web/packages/Cairo/index.html

cairoDevice 2.24 https://cran.r-project.org/web/packages/cairoDevice/index.html

calibrate 1.7.2 https://cran.r-project.org/web/packages/calibrate/index.html

calibrator 1.2-6 https://cran.r-project.org/web/packages/calibrator/index.html

callr 2.0.2 https://cran.r-project.org/web/packages/callr/index.html

car 2.1-6 https://cran.r-project.org/web/packages/car/index.html

carData 3.0-1 https://cran.r-project.org/web/packages/carData/index.html

caret 6.0-78 https://cran.r-project.org/web/packages/caret/index.html

catnet 1.15.3 https://cran.r-project.org/web/packages/catnet/index.html

caTools 1.17.1 https://cran.r-project.org/web/packages/caTools/index.html

cclust 0.6-21 https://cran.r-project.org/web/packages/cclust/index.html

cellranger 1.1.0 https://cran.r-project.org/web/packages/cellranger/index.html

ChainLadder 0.2.5 https://cran.r-

project.org/web/packages/ChainLadder/index.html

changepoint 2.2.2 https://cran.r-

project.org/web/packages/changepoint/index.html

checkmate 1.8.5 https://cran.r-project.org/web/packages/checkmate/index.html

checkpoint 0.4.3 https://cran.r-project.org/web/packages/checkpoint/index.html

choroplethrMaps 1.0.1 https://cran.r-

Package Version Link

project.org/web/packages/choroplethrMaps/index.html

chron 2.3-52 https://cran.r-project.org/web/packages/chron/index.html

circlize 0.4.3 https://cran.r-project.org/web/packages/circlize/index.html

Ckmeans.1d.dp 4.2.1 https://cran.r-

project.org/web/packages/Ckmeans.1d.dp/index.html

class 7.3-14 https://cran.r-project.org/web/packages/class/index.html

classInt 0.3-3 https://cran.r-project.org/web/packages/classInt/index.html

CLI 1.0.0 https://cran.r-project.org/web/packages/cli/index.html

ClickClust 1.1.5 https://cran.r-project.org/web/packages/ClickClust/index.html

clickstream 1.3.0 https://cran.r-project.org/web/packages/clickstream/index.html

clue 0.3-54 https://cran.r-project.org/web/packages/clue/index.html

cluster 2.0.6 https://cran.r-project.org/web/packages/cluster/index.html

clv 0.3-2.1 https://cran.r-project.org/web/packages/clv/index.html

cmprsk 2.2-7 https://cran.r-project.org/web/packages/cmprsk/index.html

coda 0.19-1 https://cran.r-project.org/web/packages/coda/index.html

codetools 0.2-15 https://cran.r-project.org/web/packages/codetools/index.html

coefplot 1.2.6 https://cran.r-project.org/web/packages/coefplot/index.html

coin 1.2-2 https://cran.r-project.org/web/packages/coin/index.html

collapsibleTree 0.1.6 https://cran.r-

project.org/web/packages/collapsibleTree/index.html

colorRamps 2.3 https://cran.r-

project.org/web/packages/colorRamps/index.html

colorspace 1.3-2 https://cran.r-project.org/web/packages/colorspace/index.html

colourpicker 1.0 https://cran.r-

project.org/web/packages/colourpicker/index.html

combinat 0.0-8 https://cran.r-project.org/web/packages/combinat/index.html

commonmark 1.4 https://cran.r-

project.org/web/packages/commonmark/index.html

compiler 3.4.4 NA
Package Version Link

compositions 1.40-1 https://cran.r-

project.org/web/packages/compositions/index.html

CORElearn 1.52.0 https://cran.r-project.org/web/packages/CORElearn/index.html

corpcor 1.6.9 https://cran.r-project.org/web/packages/corpcor/index.html

corrgram 1.12 https://cran.r-project.org/web/packages/corrgram/index.html

corrplot 0.84 https://cran.r-project.org/web/packages/corrplot/index.html

covr 3.0.1 https://cran.r-project.org/web/packages/covr/index.html

cowplot 0.9.2 https://cran.r-project.org/web/packages/cowplot/index.html

cplm 0.7-5 https://cran.r-project.org/web/packages/cplm/index.html

cpp11 0.4.2 https://cran.r-project.org/web/packages/cpp11/index.html

crayon 1.3.4 https://cran.r-project.org/web/packages/crayon/index.html

crosstalk 1.0.0 https://cran.r-project.org/web/packages/crosstalk/index.html

cslogistic 0.1-3 https://cran.r-project.org/web/packages/cslogistic/index.html

cts 1.0-21 https://cran.r-project.org/web/packages/cts/index.html

ctv 0.8-4 https://cran.r-project.org/web/packages/ctv/index.html

cubature 1.3-11 https://cran.r-project.org/web/packages/cubature/index.html

Cubist 0.2.1 https://cran.r-project.org/web/packages/Cubist/index.html

curl 3.2 https://cran.r-project.org/web/packages/curl/index.html

CVST 0.2-1 https://cran.r-project.org/web/packages/CVST/index.html

cvTools 0.3.2 https://cran.r-project.org/web/packages/cvTools/index.html

d3heatmap 0.6.1.2 https://cran.r-project.org/web/packages/d3heatmap/index.html

d3Network 0.5.2.1 https://cran.r-project.org/web/packages/d3Network/index.html

d3r 0.8.0 https://cran.r-project.org/web/packages/d3r/index.html

data.table 1.12.6 https://cran.r-project.org/web/packages/data.table/index.html

data.tree 0.7.5 https://cran.r-project.org/web/packages/data.tree/index.html

datasauRus 0.1.4 https://cran.r-project.org/web/packages/datasauRus/index.html

semantic models 3.4.4 NA

Package Version Link

date 1.2-38 https://cran.r-project.org/web/packages/date/index.html

DBI 0.8 https://cran.r-project.org/web/packages/DBI/index.html

dbplyr 1.2.1 https://cran.r-project.org/web/packages/dbplyr/index.html

dbscan 1.1-1 https://cran.r-project.org/web/packages/dbscan/index.html

dclone 2.2-0 https://cran.r-project.org/web/packages/dclone/index.html

ddalpha 1.3.1.1 https://cran.r-project.org/web/packages/ddalpha/index.html

deal 1.2-37 https://cran.r-project.org/web/packages/deal/index.html

debugme 1.1.0 https://cran.r-project.org/web/packages/debugme/index.html

deepnet 0.2 https://cran.r-project.org/web/packages/deepnet/index.html

deldir 0.1-14 https://cran.r-project.org/web/packages/deldir/index.html

dendextend 1.12.0 https://cran.r-

project.org/web/packages/dendextend/index.html

DEoptimR 1.0-8 https://cran.r-project.org/web/packages/DEoptimR/index.html

deployrRserve 9.0.0 NA

Deriv 3.8.4 https://cran.r-project.org/web/packages/Deriv/index.html

desc 1.1.1 https://cran.r-project.org/web/packages/desc/index.html

descr 1.1.4 https://cran.r-project.org/web/packages/descr/index.html

deSolve 1.20 https://cran.r-project.org/web/packages/deSolve/index.html

devtools 1.13.5 https://cran.r-project.org/web/packages/devtools/index.html

DiagrammeR 1.0.0 https://cran.r-

project.org/web/packages/DiagrammeR/index.html

DiagrammeRsvg 0.1 https://cran.r-

project.org/web/packages/DiagrammeRsvg/index.html

dichromat 2.0-0 https://cran.r-project.org/web/packages/dichromat/index.html

digest 0.6.15 https://cran.r-project.org/web/packages/digest/index.html

dimRed 0.1.0 https://cran.r-project.org/web/packages/dimRed/index.html

diptest 0.75-7 https://cran.r-project.org/web/packages/diptest/index.html

Package Version Link

distcrete 1.0.3 https://cran.r-project.org/web/packages/distcrete/index.html

DistributionUtils 0.6-0 https://cran.r-

project.org/web/packages/DistributionUtils/index.html

distrom 1.0 https://cran.r-project.org/web/packages/distrom/index.html

dlm 1.1-4 https://cran.r-project.org/web/packages/dlm/index.html

DMwR 0.4.1 https://cran.r-project.org/web/packages/DMwR/index.html

doBy 4.6-1 https://cran.r-project.org/web/packages/doBy/index.html

doParallel 1.0.12 https://cran.r-project.org/web/packages/doParallel/index.html

doSNOW 1.0.16 https://cran.r-project.org/web/packages/doSNOW/index.html

dotCall64 0.9-5.2 https://cran.r-project.org/web/packages/dotCall64/index.html

downloader 0.4 https://cran.r-

project.org/web/packages/downloader/index.html

dplyr 0.8.3 https://cran.r-project.org/web/packages/dplyr/index.html

DPpackage 1.1-7.4 https://cran.r-project.org/web/packages/DPpackage/index.html

DRR 0.0.3 https://cran.r-project.org/web/packages/DRR/index.html

dse 2015.12-1 https://cran.r-project.org/web/packages/dse/index.html

DT 0.4 https://cran.r-project.org/web/packages/DT/index.html

dtt 0.1-2 https://cran.r-project.org/web/packages/dtt/index.html

dtw 1.18-1 https://cran.r-project.org/web/packages/dtw/index.html

dygraphs 1.1.1.4 https://cran.r-project.org/web/packages/dygraphs/index.html

dynlm 0.3-5 https://cran.r-project.org/web/packages/dynlm/index.html

e1071 1.6-8 https://cran.r-project.org/web/packages/e1071/index.html

earth 4.6.2 https://cran.r-project.org/web/packages/earth/index.html

EbayesThresh 1.4-12 https://cran.r-

project.org/web/packages/EbayesThresh/index.html

ebdbNet 1.2.5 https://cran.r-project.org/web/packages/ebdbNet/index.html

ecm 4.4.0 https://cran.r-project.org/web/packages/ecm/index.html

Package Version Link

edeaR 0.8.0 https://cran.r-project.org/web/packages/edeaR/index.html

effects 4.0-1 https://cran.r-project.org/web/packages/effects/index.html

ellipse 0.4.1 https://cran.r-project.org/web/packages/ellipse/index.html

ellipsis 0.3.0 https://cran.r-project.org/web/packages/ellipsis/index.html

emmeans 1.1.2 https://cran.r-project.org/web/packages/emmeans/index.html

emulator 1.2-15 https://cran.r-project.org/web/packages/emulator/index.html

energy 1.7-2 https://cran.r-project.org/web/packages/energy/index.html

english 1.2-3 https://cran.r-project.org/web/packages/english/index.html

ensembleBMA 5.1.5 https://cran.r-

project.org/web/packages/ensembleBMA/index.html

entropy 1.2.1 https://cran.r-project.org/web/packages/entropy/index.html

epitools 0.5-10.1 https://cran.r-project.org/web/packages/epitools/index.html

epitrix 0.2.2 https://cran.r-project.org/web/packages/epitrix/index.html

estimability 1.3 https://cran.r-project.org/web/packages/estimability/index.html

eulerr 5.1.0 https://cran.r-project.org/web/packages/eulerr/index.html

EvalEst 2015.4-2 https://cran.r-project.org/web/packages/EvalEst/index.html

evaluate 0.10.1 https://cran.r-project.org/web/packages/evaluate/index.html

evd 2.3-2 https://cran.r-project.org/web/packages/evd/index.html

evdbayes 1.1-1 https://cran.r-project.org/web/packages/evdbayes/index.html

eventdataR 0.2.0 https://cran.r-project.org/web/packages/eventdataR/index.html

exactLoglinTest 1.4.2 https://cran.r-

project.org/web/packages/exactLoglinTest/index.html

exactRankTests 0.8-29 https://cran.r-

project.org/web/packages/exactRankTests/index.html

expint 0.1-4 https://cran.r-project.org/web/packages/expint/index.html

expm 0.999-2 https://cran.r-project.org/web/packages/expm/index.html

extraDistr 1.8.8 https://cran.r-project.org/web/packages/extraDistr/index.html

Package Version Link

extrafont 0.17 https://cran.r-project.org/web/packages/extrafont/index.html

extrafontdb 1.0 https://cran.r-

project.org/web/packages/extrafontdb/index.html

extremevalues 2.3.2 https://cran.r-

project.org/web/packages/extremevalues/index.html

ez 4.4-0 https://cran.r-project.org/web/packages/ez/index.html

factoextra 1.0.5 https://cran.r-project.org/web/packages/factoextra/index.html

FactoMineR 1.40 https://cran.r-

project.org/web/packages/FactoMineR/index.html

factorQR 0.1-4 https://cran.r-project.org/web/packages/factorQR/index.html

fansi 0.4.0 https://cran.r-project.org/web/packages/fansi/index.html

faoutlier 0.7.2 https://cran.r-project.org/web/packages/faoutlier/index.html

farver 1.1.0 https://cran.r-project.org/web/packages/farver/index.html

fastICA 1.2-1 https://cran.r-project.org/web/packages/fastICA/index.html

fastmatch 1.1-0 https://cran.r-project.org/web/packages/fastmatch/index.html

fBasics 3042.89 https://cran.r-project.org/web/packages/fBasics/index.html

fdrtool 1.2.15 https://cran.r-project.org/web/packages/fdrtool/index.html

fGarch 3042.83.1 https://cran.r-project.org/web/packages/fGarch/index.html

fields 9.6 https://cran.r-project.org/web/packages/fields/index.html

filehash 2.4-1 https://cran.r-project.org/web/packages/filehash/index.html

FinCal 0.6.3 https://cran.r-project.org/web/packages/FinCal/index.html

fitdistrplus 1.0-9 https://cran.r-project.org/web/packages/fitdistrplus/index.html

flashClust 1.01-2 https://cran.r-project.org/web/packages/flashClust/index.html

flexclust 1.3-5 https://cran.r-project.org/web/packages/flexclust/index.html

flexmix 2.3-14 https://cran.r-project.org/web/packages/flexmix/index.html

FME 1.3.5 https://cran.r-project.org/web/packages/FME/index.html

fmsb 0.6.1 https://cran.r-project.org/web/packages/fmsb/index.html

Package Version Link

FNN 1.1 https://cran.r-project.org/web/packages/FNN/index.html

fontBitstreamVera 0.1.1 https://cran.r-

project.org/web/packages/fontBitstreamVera/index.html

fontLiberation 0.1.0 https://cran.r-

project.org/web/packages/fontLiberation/index.html

fontquiver 0.2.1 https://cran.r-project.org/web/packages/fontquiver/index.html

forcats 0.3.0 https://cran.r-project.org/web/packages/forcats/index.html

foreach 1.4.4 https://cran.r-project.org/web/packages/foreach/index.html

forecast 8.7 https://cran.r-project.org/web/packages/forecast/index.html

forecastHybrid 2.1.11 https://cran.r-

project.org/web/packages/forecastHybrid/index.html

foreign 0.8-69 https://cran.r-project.org/web/packages/foreign/index.html

formatR 1.5 https://cran.r-project.org/web/packages/formatR/index.html

formattable 0.2.0.1 https://cran.r-

project.org/web/packages/formattable/index.html

Formula 1.2-2 https://cran.r-project.org/web/packages/Formula/index.html

fpc 2.1-11 https://cran.r-project.org/web/packages/fpc/index.html

fracdiff 1.4-2 https://cran.r-project.org/web/packages/fracdiff/index.html

fTrading 3042.79 https://cran.r-project.org/web/packages/fTrading/index.html

fUnitRoots 3042.79 https://cran.r-project.org/web/packages/fUnitRoots/index.html

futile.logger 1.4.3 https://cran.r-

project.org/web/packages/futile.logger/index.html

futile.options 1.0.0 https://cran.r-

project.org/web/packages/futile.options/index.html

future 1.15.0 https://cran.r-project.org/web/packages/future/index.html

future.apply 1.3.0 https://cran.r-

project.org/web/packages/future.apply/index.html

gam 1.15 https://cran.r-project.org/web/packages/gam/index.html

gamlr 1.13-4 https://cran.r-project.org/web/packages/gamlr/index.html

Package Version Link

gamlss 5.0-6 https://cran.r-project.org/web/packages/gamlss/index.html

gamlss.data 5.0-0 https://cran.r-

project.org/web/packages/gamlss.data/index.html

gamlss.dist 5.0-4 https://cran.r-project.org/web/packages/gamlss.dist/index.html

gbm 2.1.3 https://cran.r-project.org/web/packages/gbm/index.html

gclus 1.3.1 https://cran.r-project.org/web/packages/gclus/index.html

gdalUtils 2.0.1.7 https://cran.r-project.org/web/packages/gdalUtils/index.html

gdata 2.18.0 https://cran.r-project.org/web/packages/gdata/index.html

gdtools 0.1.7 https://cran.r-project.org/web/packages/gdtools/index.html

gee 4.13-19 https://cran.r-project.org/web/packages/gee/index.html

genalg 0.2.0 https://cran.r-project.org/web/packages/genalg/index.html

generics 0.1.2 https://cran.r-project.org/web/packages/generics/index.html

genetics 1.3.8.1 https://cran.r-project.org/web/packages/genetics/index.html

GenSA 1.1.7 https://cran.r-project.org/web/packages/GenSA/index.html

geojson 0.2.0 https://cran.r-project.org/web/packages/geojson/index.html

geojsonio 0.6.0 https://cran.r-project.org/web/packages/geojsonio/index.html

geojsonlint 0.2.0 https://cran.r-project.org/web/packages/geojsonlint/index.html

geoR 1.7-5.2 https://cran.r-project.org/web/packages/geoR/index.html

geoRglm 0.9-11 https://cran.r-project.org/web/packages/geoRglm/index.html

geosphere 1.5-7 https://cran.r-project.org/web/packages/geosphere/index.html

GGally 2.0.0 https://cran.r-project.org/web/packages/GGally/index.html

ggalt 0.4.0 https://cran.r-project.org/web/packages/ggalt/index.html

gganimate 1.0.3 https://cran.r-project.org/web/packages/gganimate/index.html

ggcorrplot 0.1.1 https://cran.r-project.org/web/packages/ggcorrplot/index.html

ggdendro 0.1-20 https://cran.r-project.org/web/packages/ggdendro/index.html

ggeffects 0.3.2 https://cran.r-project.org/web/packages/ggeffects/index.html

ggExtra 0.9 https://cran.r-project.org/web/packages/ggExtra/index.html

Package Version Link

ggforce 0.1.1 https://cran.r-project.org/web/packages/ggforce/index.html

ggformula 0.6.2 https://cran.r-project.org/web/packages/ggformula/index.html

ggfortify 0.4.3 https://cran.r-project.org/web/packages/ggfortify/index.html

gghighlight 0.3.0 https://cran.r-

project.org/web/packages/gghighlight/index.html

ggimage 0.1.2 https://cran.r-project.org/web/packages/ggimage/index.html

ggiraph 0.6.1 https://cran.r-project.org/web/packages/ggiraph/index.html

ggjoy 0.4.0 https://cran.r-project.org/web/packages/ggjoy/index.html

ggm 2.3 https://cran.r-project.org/web/packages/ggm/index.html

ggmap 3.0.0 https://cran.r-project.org/web/packages/ggmap/index.html

ggmcmc 1.1 https://cran.r-project.org/web/packages/ggmcmc/index.html

ggplot2 3.3.3 https://cran.r-project.org/web/packages/ggplot2/index.html

ggplot2movies 0.0.1 https://cran.r-

project.org/web/packages/ggplot2movies/index.html

ggpmisc 0.2.16 https://cran.r-project.org/web/packages/ggpmisc/index.html

ggpubr 0.2.3 https://cran.r-project.org/web/packages/ggpubr/index.html

ggQC 0.0.31 https://cran.r-project.org/web/packages/ggQC/index.html

ggRandomForests 2.0.1 https://cran.r-

project.org/web/packages/ggRandomForests/index.html

ggraph 1.0.1 https://cran.r-project.org/web/packages/ggraph/index.html

ggrepel 0.8.0 https://cran.r-project.org/web/packages/ggrepel/index.html

ggridges 0.4.1 https://cran.r-project.org/web/packages/ggridges/index.html

ggsci 2.8 https://cran.r-project.org/web/packages/ggsci/index.html

ggsignif 0.4.0 https://cran.r-project.org/web/packages/ggsignif/index.html

ggsoccer 0.1.4 https://cran.r-project.org/web/packages/ggsoccer/index.html

ggstance 0.3 https://cran.r-project.org/web/packages/ggstance/index.html

ggtern 2.2.1 https://cran.r-project.org/web/packages/ggtern/index.html

Package Version Link

ggthemes 3.4.0 https://cran.r-project.org/web/packages/ggthemes/index.html

gistr 0.4.0 https://cran.r-project.org/web/packages/gistr/index.html

git2r 0.21.0 https://cran.r-project.org/web/packages/git2r/index.html

glasso 1.8 https://cran.r-project.org/web/packages/glasso/index.html

glmmBUGS 2.4.0 https://cran.r-project.org/web/packages/glmmBUGS/index.html

glmmTMB 0.2.0 https://cran.r-project.org/web/packages/glmmTMB/index.html

glmnet 2.0-13 https://cran.r-project.org/web/packages/glmnet/index.html

GlobalOptions 0.0.13 https://cran.r-

project.org/web/packages/GlobalOptions/index.html

globals 0.12.4 https://cran.r-project.org/web/packages/globals/index.html

glue 1.3.1 https://cran.r-project.org/web/packages/glue/index.html

gmodels 2.16.2 https://cran.r-project.org/web/packages/gmodels/index.html

gmp 0.5-13.1 https://cran.r-project.org/web/packages/gmp/index.html

gnm 1.0-8 https://cran.r-project.org/web/packages/gnm/index.html

goftest 1.1-1 https://cran.r-project.org/web/packages/goftest/index.html

googleVis 0.6.2 https://cran.r-project.org/web/packages/googleVis/index.html

gower 0.1.2 https://cran.r-project.org/web/packages/gower/index.html

GPArotation 2014.11-1 https://cran.r-

project.org/web/packages/GPArotation/index.html

gplots 3.0.1 https://cran.r-project.org/web/packages/gplots/index.html

graphics 3.4.4 NA

grDevices 3.4.4 NA

grid 3.4.4 NA

gridBase 0.4-7 https://cran.r-project.org/web/packages/gridBase/index.html

gridExtra 2.3 https://cran.r-project.org/web/packages/gridExtra/index.html

gridGraphics 0.2-1 https://cran.r-

project.org/web/packages/gridGraphics/index.html
Package Version Link

growcurves 0.2.4.1 https://cran.r-project.org/web/packages/growcurves/index.html

grpreg 3.1-2 https://cran.r-project.org/web/packages/grpreg/index.html

gss 2.1-7 https://cran.r-project.org/web/packages/gss/index.html

gstat 1.1-5 https://cran.r-project.org/web/packages/gstat/index.html

gsubfn 0.7 https://cran.r-project.org/web/packages/gsubfn/index.html

gtable 0.2.0 https://cran.r-project.org/web/packages/gtable/index.html

gtools 3.5.0 https://cran.r-project.org/web/packages/gtools/index.html

gtrendsR 1.4.3 https://cran.r-project.org/web/packages/gtrendsR/index.html

gWidgets 0.0-54 https://cran.r-project.org/web/packages/gWidgets/index.html

gWidgetsRGtk2 0.0-86 https://cran.r-

project.org/web/packages/gWidgetsRGtk2/index.html

gWidgetstcltk 0.0-55 https://cran.r-

project.org/web/packages/gWidgetstcltk/index.html

haplo.stats 1.7.7 https://cran.r-project.org/web/packages/haplo.stats/index.html

hash 2.2.6 https://cran.r-project.org/web/packages/hash/index.html

haven 1.1.1 https://cran.r-project.org/web/packages/haven/index.html

hbsae 1.0 https://cran.r-project.org/web/packages/hbsae/index.html

HDInterval 0.2.0 https://cran.r-project.org/web/packages/HDInterval/index.html

hdrcde 3.2 https://cran.r-project.org/web/packages/hdrcde/index.html

heatmaply 0.16.0 https://cran.r-project.org/web/packages/heatmaply/index.html

heavy 0.38.19 https://cran.r-project.org/web/packages/heavy/index.html

hexbin 1.27.2 https://cran.r-project.org/web/packages/hexbin/index.html

hflights 0.1 https://cran.r-project.org/web/packages/hflights/index.html

HH 3.1-34 https://cran.r-project.org/web/packages/HH/index.html

HI 0.4 https://cran.r-project.org/web/packages/HI/index.html

highcharter 0.5.0 https://cran.r-project.org/web/packages/highcharter/index.html

highr 0.6 https://cran.r-project.org/web/packages/highr/index.html

Package Version Link

HistData 0.8-2 https://cran.r-project.org/web/packages/HistData/index.html

Hmisc 4.1-1 https://cran.r-project.org/web/packages/Hmisc/index.html

hms 0.4.2 https://cran.r-project.org/web/packages/hms/index.html

hoardr 0.2.0 https://cran.r-project.org/web/packages/hoardr/index.html

hrbrthemes 0.6.0 https://cran.r-project.org/web/packages/hrbrthemes/index.html

HSAUR 1.3-9 https://cran.r-project.org/web/packages/HSAUR/index.html

htmlTable 1.11.2 https://cran.r-project.org/web/packages/htmlTable/index.html

htmltools 0.3.6 https://cran.r-project.org/web/packages/htmltools/index.html

htmlwidgets 1.3 https://cran.r-

project.org/web/packages/htmlwidgets/index.html

hts 5.1.5 https://cran.r-project.org/web/packages/hts/index.html

httpuv 1.3.6.2 https://cran.r-project.org/web/packages/httpuv/index.html

httr 1.3.1 https://cran.r-project.org/web/packages/httr/index.html

huge 1.2.7 https://cran.r-project.org/web/packages/huge/index.html

hunspell 2.9 https://cran.r-project.org/web/packages/hunspell/index.html

hydroTSM 0.5-1 https://cran.r-project.org/web/packages/hydroTSM/index.html

IBrokers 0.9-12 https://cran.r-project.org/web/packages/IBrokers/index.html

ifultools 2.0-4 https://cran.r-project.org/web/packages/ifultools/index.html

igraph 1.2.1 https://cran.r-project.org/web/packages/igraph/index.html

imager 0.40.2 https://cran.r-project.org/web/packages/imager/index.html

imputeTS 2.7 https://cran.r-project.org/web/packages/imputeTS/index.html

incidence 1.7.2 https://cran.r-project.org/web/packages/incidence/index.html

influenceR 0.1.0 https://cran.r-project.org/web/packages/influenceR/index.html

InformationValue 1.2.3 https://cran.r-

project.org/web/packages/InformationValue/index.html

inline 0.3.14 https://cran.r-project.org/web/packages/inline/index.html

intervals 0.15.1 https://cran.r-project.org/web/packages/intervals/index.html

Package Version Link

inum 1.0-0 https://cran.r-project.org/web/packages/inum/index.html

investr 1.4.2 https://cran.r-project.org/web/packages/investr/index.html

ipred 0.9-6 https://cran.r-project.org/web/packages/ipred/index.html

irlba 2.3.2 https://cran.r-project.org/web/packages/irlba/index.html

irr 0.84 https://cran.r-project.org/web/packages/irr/index.html

isoband 0.2.0 https://cran.r-project.org/web/packages/isoband/index.html

ISOcodes 2017.09.27 https://cran.r-project.org/web/packages/ISOcodes/index.html

iterators 1.0.9 https://cran.r-project.org/web/packages/iterators/index.html

janeaustenr 0.1.5 https://cran.r-

project.org/web/packages/janeaustenr/index.html

janitor 1.0.0 https://cran.r-project.org/web/packages/janitor/index.html

jmvcore 1.0.8 https://cran.r-project.org/web/packages/jmvcore/index.html

jpeg 0.1-8 https://cran.r-project.org/web/packages/jpeg/index.html

jqr 1.0.0 https://cran.r-project.org/web/packages/jqr/index.html

jsonlite 1.6 https://cran.r-project.org/web/packages/jsonlite/index.html

jsonvalidate 1.0.0 https://cran.r-

project.org/web/packages/jsonvalidate/index.html

jtools 0.9.4 https://cran.r-project.org/web/packages/jtools/index.html

kableExtra 0.7.0 https://cran.r-project.org/web/packages/kableExtra/index.html

Kendall 2.2 https://cran.r-project.org/web/packages/Kendall/index.html

kernlab 0.9-25 https://cran.r-project.org/web/packages/kernlab/index.html

KernSmooth 2.23-15 https://cran.r-

project.org/web/packages/KernSmooth/index.html

KFKSDS 1.6 https://cran.r-project.org/web/packages/KFKSDS/index.html

kinship2 1.6.4 https://cran.r-project.org/web/packages/kinship2/index.html

kknn 1.3.1 https://cran.r-project.org/web/packages/kknn/index.html

klaR 0.6-14 https://cran.r-project.org/web/packages/klaR/index.html

Package Version Link

km.ci 0.5-2 https://cran.r-project.org/web/packages/km.ci/index.html

KMsurv 0.1-5 https://cran.r-project.org/web/packages/KMsurv/index.html

knitr 1.20 https://cran.r-project.org/web/packages/knitr/index.html

ks 1.11.0 https://cran.r-project.org/web/packages/ks/index.html

labeling 0.3 https://cran.r-project.org/web/packages/labeling/index.html

labelled 1.0.1 https://cran.r-project.org/web/packages/labelled/index.html

laeken 0.4.6 https://cran.r-project.org/web/packages/laeken/index.html

Lahman 6.0-0 https://cran.r-project.org/web/packages/Lahman/index.html

lambda.r 1.2 https://cran.r-project.org/web/packages/lambda.r/index.html

lars 1.2 https://cran.r-project.org/web/packages/lars/index.html

later 1.0.0 https://cran.r-project.org/web/packages/later/index.html

latex2exp 0.4.0 https://cran.r-project.org/web/packages/latex2exp/index.html

lattice 0.20-35 https://cran.r-project.org/web/packages/lattice/index.html

latticeExtra 0.6-28 https://cran.r-project.org/web/packages/latticeExtra/index.html

lava 1.6.1 https://cran.r-project.org/web/packages/lava/index.html

lavaan 0.5- https://cran.r-project.org/web/packages/lavaan/index.html

23.1097

lazyeval 0.2.1 https://cran.r-project.org/web/packages/lazyeval/index.html

lda 1.4.2 https://cran.r-project.org/web/packages/lda/index.html

leaflet 2.0.2 https://cran.r-project.org/web/packages/leaflet/index.html

leaflet.esri 0.2 https://cran.r-project.org/web/packages/leaflet.esri/index.html

leaflet.extras 0.2 https://cran.r-

project.org/web/packages/leaflet.extras/index.html

leaps 3.0 https://cran.r-project.org/web/packages/leaps/index.html

LearnBayes 2.15.1 https://cran.r-project.org/web/packages/LearnBayes/index.html

lexicon 1.2.1 https://cran.r-project.org/web/packages/lexicon/index.html

libcoin 1.0-1 https://cran.r-project.org/web/packages/libcoin/index.html

Package Version Link

LiblineaR 2.10-8 https://cran.r-project.org/web/packages/LiblineaR/index.html

LICORS 0.2.0 https://cran.r-project.org/web/packages/LICORS/index.html

lifecycle 0.1.0 https://cran.r-project.org/web/packages/lifecycle/index.html

likert 1.3.5 https://cran.r-project.org/web/packages/likert/index.html

limSolve 1.5.5.3 https://cran.r-project.org/web/packages/limSolve/index.html

linelist 0.0.40.9000 https://cran.r-project.org/web/packages/linelist/index.html

linprog 0.9-2 https://cran.r-project.org/web/packages/linprog/index.html

listenv 0.7.0 https://cran.r-project.org/web/packages/listenv/index.html

lm.beta 1.5-1 https://cran.r-project.org/web/packages/lm.beta/index.html

lme4 1.1-16 https://cran.r-project.org/web/packages/lme4/index.html

lmm 1.0 https://cran.r-project.org/web/packages/lmm/index.html

lmtest 0.9-35 https://cran.r-project.org/web/packages/lmtest/index.html

locfit 1.5-9.1 https://cran.r-project.org/web/packages/locfit/index.html

locpol 0.6-0 https://cran.r-project.org/web/packages/locpol/index.html

LogicReg 1.5.9 https://cran.r-project.org/web/packages/LogicReg/index.html

lpSolve 5.6.13 https://cran.r-project.org/web/packages/lpSolve/index.html

lsa 0.73.1 https://cran.r-project.org/web/packages/lsa/index.html

lsmeans 2.27-61 https://cran.r-project.org/web/packages/lsmeans/index.html

lubridate 1.7.2 https://cran.r-project.org/web/packages/lubridate/index.html

magic 1.5-8 https://cran.r-project.org/web/packages/magic/index.html

magick 1.8 https://cran.r-project.org/web/packages/magick/index.html

magrittr 1.5 https://cran.r-project.org/web/packages/magrittr/index.html

manipulateWidget 0.9.0 https://cran.r-

project.org/web/packages/manipulateWidget/index.html

MAPA 2.0.4 https://cran.r-project.org/web/packages/MAPA/index.html

mapdata 2.3.0 https://cran.r-project.org/web/packages/mapdata/index.html

mapproj 1.2.6 https://cran.r-project.org/web/packages/mapproj/index.html

Package Version Link

maps 3.2.0 https://cran.r-project.org/web/packages/maps/index.html

maptools 0.9-2 https://cran.r-project.org/web/packages/maptools/index.html

maptree 1.4-7 https://cran.r-project.org/web/packages/maptree/index.html

mapview 2.3.0 https://cran.r-project.org/web/packages/mapview/index.html

marima 2.2 https://cran.r-project.org/web/packages/marima/index.html

markdown 0.8 https://cran.r-project.org/web/packages/markdown/index.html

MASS 7.3-49 https://cran.r-project.org/web/packages/MASS/index.html

MasterBayes 2.55 https://cran.r-

project.org/web/packages/MasterBayes/index.html

Matching 4.9-5 https://cran.r-project.org/web/packages/Matching/index.html

MatchIt 3.0.2 https://cran.r-project.org/web/packages/MatchIt/index.html

matchmaker 0.1.1 https://cran.r-

project.org/web/packages/matchmaker/index.html

Matrix 1.2-12 https://cran.r-project.org/web/packages/Matrix/index.html

matrixcalc 1.0-3 https://cran.r-project.org/web/packages/matrixcalc/index.html

MatrixModels 0.4-1 https://cran.r-

project.org/web/packages/MatrixModels/index.html

matrixStats 0.54.0 https://cran.r-project.org/web/packages/matrixStats/index.html

maxent 1.3.3.1 https://cran.r-project.org/web/packages/maxent/index.html

maxLik 1.3-4 https://cran.r-project.org/web/packages/maxLik/index.html

maxstat 0.7-25 https://cran.r-project.org/web/packages/maxstat/index.html

mboost 2.8-1 https://cran.r-project.org/web/packages/mboost/index.html

mclust 5.4 https://cran.r-project.org/web/packages/mclust/index.html

mcmc 0.9-5 https://cran.r-project.org/web/packages/mcmc/index.html

MCMCglmm 2.25 https://cran.r-

project.org/web/packages/MCMCglmm/index.html

mda 0.4-10 https://cran.r-project.org/web/packages/mda/index.html

memoise 1.1.0 https://cran.r-project.org/web/packages/memoise/index.html

Package Version Link

merTools 0.3.0 https://cran.r-project.org/web/packages/merTools/index.html

meta 4.9-1 https://cran.r-project.org/web/packages/meta/index.html

metafor 2.0-0 https://cran.r-project.org/web/packages/metafor/index.html

methods 3.4.4 NA

metricsgraphics 0.9.0 https://cran.r-

project.org/web/packages/metricsgraphics/index.html

mgcv 1.8-23 https://cran.r-project.org/web/packages/mgcv/index.html

mgsub 1.7.1 https://cran.r-project.org/web/packages/mgsub/index.html

mi 1.0 https://cran.r-project.org/web/packages/mi/index.html

mice 2.46.0 https://cran.r-project.org/web/packages/mice/index.html

microbenchmark 1.4-4 https://cran.r-

project.org/web/packages/microbenchmark/index.html

MicrosoftR 3.4.4.0105 NA

mime 0.5 https://cran.r-project.org/web/packages/mime/index.html

miniCRAN 0.2.11 https://cran.r-project.org/web/packages/miniCRAN/index.html

miniUI 0.1.1 https://cran.r-project.org/web/packages/miniUI/index.html

minpack.lm 1.2-1 https://cran.r-project.org/web/packages/minpack.lm/index.html

minqa 1.2.4 https://cran.r-project.org/web/packages/minqa/index.html

mirt 1.27.1 https://cran.r-project.org/web/packages/mirt/index.html

misc3d 0.8-4 https://cran.r-project.org/web/packages/misc3d/index.html

miscTools 0.6-22 https://cran.r-project.org/web/packages/miscTools/index.html

mitools 2.3 https://cran.r-project.org/web/packages/mitools/index.html

mixtools 1.1.0 https://cran.r-project.org/web/packages/mixtools/index.html

mlapi 0.1.0 https://cran.r-project.org/web/packages/mlapi/index.html

mlbench 2.1-1 https://cran.r-project.org/web/packages/mlbench/index.html

mlogitBMA 0.1-6 https://cran.r-project.org/web/packages/mlogitBMA/index.html

mnormt 1.5-5 https://cran.r-project.org/web/packages/mnormt/index.html

Package Version Link

MNP 3.1-0 https://cran.r-project.org/web/packages/MNP/index.html

ModelMetrics 1.1.0 https://cran.r-

project.org/web/packages/ModelMetrics/index.html

modelr 0.1.1 https://cran.r-project.org/web/packages/modelr/index.html

modeltools 0.2-21 https://cran.r-project.org/web/packages/modeltools/index.html

mombf 1.9.6 https://cran.r-project.org/web/packages/mombf/index.html

moments 0.14 https://cran.r-project.org/web/packages/moments/index.html

monomvn 1.9-7 https://cran.r-project.org/web/packages/monomvn/index.html

monreg 0.1.3 https://cran.r-project.org/web/packages/monreg/index.html

mosaic 1.1.1 https://cran.r-project.org/web/packages/mosaic/index.html

mosaicCore 0.4.2 https://cran.r-

project.org/web/packages/mosaicCore/index.html

mosaicData 0.16.0 https://cran.r-

project.org/web/packages/mosaicData/index.html

MSBVAR 0.9-3 https://cran.r-project.org/web/packages/MSBVAR/index.html

msir 1.3.2 https://cran.r-project.org/web/packages/msir/index.html

msm 1.6.6 https://cran.r-project.org/web/packages/msm/index.html

multcomp 1.4-8 https://cran.r-project.org/web/packages/multcomp/index.html

multicool 0.1-10 https://cran.r-project.org/web/packages/multicool/index.html

munsell 0.5.0 https://cran.r-project.org/web/packages/munsell/index.html

mvoutlier 2.0.9 https://cran.r-project.org/web/packages/mvoutlier/index.html

mvtnorm 1.0-7 https://cran.r-project.org/web/packages/mvtnorm/index.html

NbClust 3.0 https://cran.r-project.org/web/packages/NbClust/index.html

ncvreg 3.9-1 https://cran.r-project.org/web/packages/ncvreg/index.html

network 1.13.0 https://cran.r-project.org/web/packages/network/index.html

networkD3 0.4 https://cran.r-project.org/web/packages/networkD3/index.html

neuralnet 1.33 https://cran.r-project.org/web/packages/neuralnet/index.html

Package Version Link

ngram 3.0.4 https://cran.r-project.org/web/packages/ngram/index.html

nlme 3.1-131.1 https://cran.r-project.org/web/packages/nlme/index.html

nloptr 1.0.4 https://cran.r-project.org/web/packages/nloptr/index.html

NLP 0.1-11 https://cran.r-project.org/web/packages/NLP/index.html

nls.multstart 1.2.0 https://cran.r-

project.org/web/packages/nls.multstart/index.html

NMF 0.21.0 https://cran.r-project.org/web/packages/NMF/index.html

nnet 7.3-12 https://cran.r-project.org/web/packages/nnet/index.html

nnls 1.4 https://cran.r-project.org/web/packages/nnls/index.html

nortest 1.0-4 https://cran.r-project.org/web/packages/nortest/index.html

numbers 0.6-6 https://cran.r-project.org/web/packages/numbers/index.html

numDeriv 2016.8-1 https://cran.r-project.org/web/packages/numDeriv/index.html

numform 0.4.0 https://cran.r-project.org/web/packages/numform/index.html

OceanView 1.0.4 https://cran.r-project.org/web/packages/OceanView/index.html

openair 2.3-0 https://cran.r-project.org/web/packages/openair/index.html

openssl 1.0.1 https://cran.r-project.org/web/packages/openssl/index.html

osmar 1.1-7 https://cran.r-project.org/web/packages/osmar/index.html

outbreaks 1.5.0 https://cran.r-project.org/web/packages/outbreaks/index.html

OutlierDC 0.3-0 https://cran.r-project.org/web/packages/OutlierDC/index.html

OutlierDM 1.1.1 https://cran.r-project.org/web/packages/OutlierDM/index.html

outliers 0.14 https://cran.r-project.org/web/packages/outliers/index.html

pacbpred 0.92.2 https://cran.r-project.org/web/packages/pacbpred/index.html

packcircles 0.3.3 https://cran.r-project.org/web/packages/packcircles/index.html

padr 0.4.0 https://cran.r-project.org/web/packages/padr/index.html

parallel 3.4.4 NA

partitions 1.9-19 https://cran.r-project.org/web/packages/partitions/index.html

party 1.2-4 https://cran.r-project.org/web/packages/party/index.html

Package Version Link

partykit 1.2-0 https://cran.r-project.org/web/packages/partykit/index.html

PAWL 0.5 https://cran.r-project.org/web/packages/PAWL/index.html

pbapply 1.3-4 https://cran.r-project.org/web/packages/pbapply/index.html

pbivnorm 0.6.0 https://cran.r-project.org/web/packages/pbivnorm/index.html

pbkrtest 0.4-7 https://cran.r-project.org/web/packages/pbkrtest/index.html

PCAmixdata 3.1 https://cran.r-

project.org/web/packages/PCAmixdata/index.html

pcaPP 1.9-73 https://cran.r-project.org/web/packages/pcaPP/index.html

pdc 1.0.3 https://cran.r-project.org/web/packages/pdc/index.html

pegas 0.12 https://cran.r-project.org/web/packages/pegas/index.html

PerformanceAnalytics 1.5.2 https://cran.r-

project.org/web/packages/PerformanceAnalytics/index.html

permute 0.9-4 https://cran.r-project.org/web/packages/permute/index.html

perry 0.2.0 https://cran.r-project.org/web/packages/perry/index.html

petrinetR 0.1.0 https://cran.r-project.org/web/packages/petrinetR/index.html

pheatmap 1.0.8 https://cran.r-project.org/web/packages/pheatmap/index.html

pillar 1.4.2 https://cran.r-project.org/web/packages/pillar/index.html

pixmap 0.4-11 https://cran.r-project.org/web/packages/pixmap/index.html

pkgconfig 2.0.2 https://cran.r-project.org/web/packages/pkgconfig/index.html

pkgmaker 0.22 https://cran.r-project.org/web/packages/pkgmaker/index.html

platetools 0.1.0 https://cran.r-project.org/web/packages/platetools/index.html

plogr 0.2.0 https://cran.r-project.org/web/packages/plogr/index.html

plot3D 1.1.1 https://cran.r-project.org/web/packages/plot3D/index.html

plot3Drgl 1.0.1 https://cran.r-project.org/web/packages/plot3Drgl/index.html

plotly 4.9.2.2 https://cran.r-project.org/web/packages/plotly/index.html

plotmo 3.3.6 https://cran.r-project.org/web/packages/plotmo/index.html

plotrix 3.7 https://cran.r-project.org/web/packages/plotrix/index.html

Package Version Link

pls 2.6-0 https://cran.r-project.org/web/packages/pls/index.html

plyr 1.8.4 https://cran.r-project.org/web/packages/plyr/index.html

png 0.1-7 https://cran.r-project.org/web/packages/png/index.html

polspline 1.1.12 https://cran.r-project.org/web/packages/polspline/index.html

polyclip 1.6-1 https://cran.r-project.org/web/packages/polyclip/index.html

polylabelr 0.1.0 https://cran.r-project.org/web/packages/polylabelr/index.html

polynom 1.3-9 https://cran.r-project.org/web/packages/polynom/index.html

ppcor 1.1 https://cran.r-project.org/web/packages/ppcor/index.html

prabclus 2.2-6 https://cran.r-project.org/web/packages/prabclus/index.html

pracma 2.1.4 https://cran.r-project.org/web/packages/pracma/index.html

praise 1.0.0 https://cran.r-project.org/web/packages/praise/index.html

precrec 0.10.1 https://cran.r-project.org/web/packages/precrec/index.html

prediction 0.2.0 https://cran.r-project.org/web/packages/prediction/index.html

predmixcor 1.1-1 https://cran.r-project.org/web/packages/predmixcor/index.html

PresenceAbsence 1.1.9 https://cran.r-

project.org/web/packages/PresenceAbsence/index.html

prettyunits 1.0.2 https://cran.r-project.org/web/packages/prettyunits/index.html

pROC 1.11.0 https://cran.r-project.org/web/packages/pROC/index.html

processmapR 0.3.3 https://cran.r-

project.org/web/packages/processmapR/index.html

processmonitR 0.1.0 https://cran.r-

project.org/web/packages/processmonitR/index.html

processx 2.0.0.1 https://cran.r-project.org/web/packages/processx/index.html

prodlim 1.6.1 https://cran.r-project.org/web/packages/prodlim/index.html

profdpm 3.3 https://cran.r-project.org/web/packages/profdpm/index.html

profileModel 0.5-9 https://cran.r-

project.org/web/packages/profileModel/index.html

progress 1.1.2 https://cran.r-project.org/web/packages/progress/index.html

Package Version Link

proj4 1.0-8 https://cran.r-project.org/web/packages/proj4/index.html

promises 1.1.0 https://cran.r-project.org/web/packages/promises/index.html

prophet 0.2.1 https://cran.r-project.org/web/packages/prophet/index.html

proto 1.0.0 https://cran.r-project.org/web/packages/proto/index.html

protolite 1.7 https://cran.r-project.org/web/packages/protolite/index.html

proxy 0.4-21 https://cran.r-project.org/web/packages/proxy/index.html

pryr 0.1.4 https://cran.r-project.org/web/packages/pryr/index.html

pscl 1.5.2 https://cran.r-project.org/web/packages/pscl/index.html

psych 1.8.3.3 https://cran.r-project.org/web/packages/psych/index.html

purrr 0.3.3 https://cran.r-project.org/web/packages/purrr/index.html

pwr 1.2-2 https://cran.r-project.org/web/packages/pwr/index.html

qap 0.1-1 https://cran.r-project.org/web/packages/qap/index.html

qcc 2.7 https://cran.r-project.org/web/packages/qcc/index.html

qdapDictionaries 1.0.7 https://cran.r-

project.org/web/packages/qdapDictionaries/index.html

qdapRegex 0.7.2 https://cran.r-project.org/web/packages/qdapRegex/index.html

qdapTools 1.3.3 https://cran.r-project.org/web/packages/qdapTools/index.html

qgraph 1.4.4 https://cran.r-project.org/web/packages/qgraph/index.html

qicharts 0.5.5 https://cran.r-project.org/web/packages/qicharts/index.html

qicharts2 0.6.0 https://cran.r-project.org/web/packages/qicharts2/index.html

quadprog 1.5-5 https://cran.r-project.org/web/packages/quadprog/index.html

qualityTools 1.55 https://cran.r-

project.org/web/packages/qualityTools/index.html

quanteda 1.1.1 https://cran.r-project.org/web/packages/quanteda/index.html

quantmod 0.4-12 https://cran.r-project.org/web/packages/quantmod/index.html

quantreg 5.35 https://cran.r-project.org/web/packages/quantreg/index.html

questionr 0.6.2 https://cran.r-project.org/web/packages/questionr/index.html

Package Version Link

qvcalc 0.9-1 https://cran.r-project.org/web/packages/qvcalc/index.html

R.matlab 3.6.1 https://cran.r-project.org/web/packages/R.matlab/index.html

R.methodsS3 1.7.1 https://cran.r-

project.org/web/packages/R.methodsS3/index.html

R.oo 1.21.0 https://cran.r-project.org/web/packages/R.oo/index.html

R.utils 2.6.0 https://cran.r-project.org/web/packages/R.utils/index.html

r2d3 0.2.3 https://cran.r-project.org/web/packages/r2d3/index.html

R2HTML 2.3.2 https://cran.r-project.org/web/packages/R2HTML/index.html

R2jags 0.5-7 https://cran.r-project.org/web/packages/R2jags/index.html

R2OpenBUGS 3.2-3.2 https://cran.r-

project.org/web/packages/R2OpenBUGS/index.html

R2WinBUGS 2.1-21 https://cran.r-

project.org/web/packages/R2WinBUGS/index.html

R6 2.2.2 https://cran.r-project.org/web/packages/R6/index.html

ramps 0.6-15 https://cran.r-project.org/web/packages/ramps/index.html

RandomFields 3.1.50 https://cran.r-

project.org/web/packages/RandomFields/index.html

RandomFieldsUtils 0.3.25 https://cran.r-

project.org/web/packages/RandomFieldsUtils/index.html

randomForest 4.6-14 https://cran.r-

project.org/web/packages/randomForest/index.html

ranger 0.9.0 https://cran.r-project.org/web/packages/ranger/index.html

RApiDatetime 0.0.4 https://cran.r-

project.org/web/packages/RApiDatetime/index.html

rappdirs 0.3.1 https://cran.r-project.org/web/packages/rappdirs/index.html

RArcInfo 0.4-12 https://cran.r-project.org/web/packages/RArcInfo/index.html

raster 2.6-7 https://cran.r-project.org/web/packages/raster/index.html

rattle 5.1.0 https://cran.r-project.org/web/packages/rattle/index.html

rayshader 0.10.1 https://cran.r-project.org/web/packages/rayshader/index.html

Package Version Link

rbenchmark 1.0.0 https://cran.r-

project.org/web/packages/rbenchmark/index.html

Rblpapi 0.3.8 https://cran.r-project.org/web/packages/Rblpapi/index.html

rbokeh 0.5.0 https://cran.r-project.org/web/packages/rbokeh/index.html

rbugs 0.5-9 https://cran.r-project.org/web/packages/rbugs/index.html

RColorBrewer 1.1-2 https://cran.r-

project.org/web/packages/RColorBrewer/index.html

Rcpp 1.0.1 https://cran.r-project.org/web/packages/Rcpp/index.html

RcppArmadillo 0.8.400.0.0 https://cran.r-

project.org/web/packages/RcppArmadillo/index.html

rcppbugs 0.1.4.2 https://cran.r-project.org/web/packages/rcppbugs/index.html

RcppDE 0.1.5 https://cran.r-project.org/web/packages/RcppDE/index.html

RcppEigen 0.3.3.4.0 https://cran.r-project.org/web/packages/RcppEigen/index.html

RcppExamples 0.1.8 https://cran.r-

project.org/web/packages/RcppExamples/index.html

RcppParallel 4.4.0 https://cran.r-

project.org/web/packages/RcppParallel/index.html

RcppProgress 0.4 https://cran.r-

project.org/web/packages/RcppProgress/index.html

RcppRoll 0.2.2 https://cran.r-project.org/web/packages/RcppRoll/index.html

RCurl 1.95-4.10 https://cran.r-project.org/web/packages/RCurl/index.html

readbitmap 0.1-4 https://cran.r-project.org/web/packages/readbitmap/index.html

readr 1.1.1 https://cran.r-project.org/web/packages/readr/index.html

readxl 1.0.0 https://cran.r-project.org/web/packages/readxl/index.html

recipes 0.1.2 https://cran.r-project.org/web/packages/recipes/index.html

Redmonder 0.2.0 https://cran.r-

project.org/web/packages/Redmonder/index.html

registry 0.5 https://cran.r-project.org/web/packages/registry/index.html

relaimpo 2.2-3 https://cran.r-project.org/web/packages/relaimpo/index.html

Package Version Link

relimp 1.0-5 https://cran.r-project.org/web/packages/relimp/index.html

rematch 1.0.1 https://cran.r-project.org/web/packages/rematch/index.html

Renext 3.1-0 https://cran.r-project.org/web/packages/Renext/index.html

reports 0.1.4 https://cran.r-project.org/web/packages/reports/index.html

reprex 0.1.2 https://cran.r-project.org/web/packages/reprex/index.html

reshape 0.8.7 https://cran.r-project.org/web/packages/reshape/index.html

reshape2 1.4.3 https://cran.r-project.org/web/packages/reshape2/index.html

reticulate 1.6 https://cran.r-project.org/web/packages/reticulate/index.html

RevoIOQ 8.0.10 NA

RevoMods 11.0.0 NA

RevoUtils 10.0.9 NA

RevoUtilsMath 10.0.1 NA

rex 1.1.2 https://cran.r-project.org/web/packages/rex/index.html

rFerns 2.0.3 https://cran.r-project.org/web/packages/rFerns/index.html

rfm 0.2.0 https://cran.r-project.org/web/packages/rfm/index.html

RGA 0.4.2 https://cran.r-project.org/web/packages/RGA/index.html

rgdal 1.2-18 https://cran.r-project.org/web/packages/rgdal/index.html

rgeos 0.3-26 https://cran.r-project.org/web/packages/rgeos/index.html

rgexf 0.15.3 https://cran.r-project.org/web/packages/rgexf/index.html

rgl 0.99.16 https://cran.r-project.org/web/packages/rgl/index.html

RgoogleMaps 1.4.1 https://cran.r-

project.org/web/packages/RgoogleMaps/index.html

RGraphics 2.0-14 https://cran.r-project.org/web/packages/RGraphics/index.html

RGtk2 2.20.34 https://cran.r-project.org/web/packages/RGtk2/index.html

RInside 0.2.14 https://cran.r-project.org/web/packages/RInside/index.html

RJaCGH 2.0.4 https://cran.r-project.org/web/packages/RJaCGH/index.html

rjags 4-6 https://cran.r-project.org/web/packages/rjags/index.html

Package Version Link

rjson 0.2.15 https://cran.r-project.org/web/packages/rjson/index.html

RJSONIO 1.3-0 https://cran.r-project.org/web/packages/RJSONIO/index.html

rlang 0.4.1 https://cran.r-project.org/web/packages/rlang/index.html

rlecuyer 0.3-4 https://cran.r-project.org/web/packages/rlecuyer/index.html

rlist 0.4.6.1 https://cran.r-project.org/web/packages/rlist/index.html

rmapshaper 0.3.0 https://cran.r-

project.org/web/packages/rmapshaper/index.html

rmarkdown 1.9 https://cran.r-project.org/web/packages/rmarkdown/index.html

Rmisc 1.5 https://cran.r-project.org/web/packages/Rmisc/index.html

Rmpfr 0.7-0 https://cran.r-project.org/web/packages/Rmpfr/index.html

rms 5.1-2 https://cran.r-project.org/web/packages/rms/index.html

RMySQL 0.10.14 https://cran.r-project.org/web/packages/RMySQL/index.html

rngtools 1.2.4 https://cran.r-project.org/web/packages/rngtools/index.html

robCompositions 2.0.6 https://cran.r-

project.org/web/packages/robCompositions/index.html

robfilter 4.1 https://cran.r-project.org/web/packages/robfilter/index.html

robustbase 0.92-8 https://cran.r-project.org/web/packages/robustbase/index.html

robustHD 0.5.1 https://cran.r-project.org/web/packages/robustHD/index.html

ROCR 1.0-7 https://cran.r-project.org/web/packages/ROCR/index.html

RODBC 1.3-15 https://cran.r-project.org/web/packages/RODBC/index.html

Rook 1.1-1 https://cran.r-project.org/web/packages/Rook/index.html

rootSolve 1.7 https://cran.r-project.org/web/packages/rootSolve/index.html

roxygen2 6.0.1 https://cran.r-project.org/web/packages/roxygen2/index.html

rpart 4.1-13 https://cran.r-project.org/web/packages/rpart/index.html

rpart.plot 2.1.2 https://cran.r-project.org/web/packages/rpart.plot/index.html

rpivotTable 0.3.0 https://cran.r-project.org/web/packages/rpivotTable/index.html

rprojroot 1.3-2 https://cran.r-project.org/web/packages/rprojroot/index.html

Package Version Link

rrcov 1.4-3 https://cran.r-project.org/web/packages/rrcov/index.html

rscproxy 2.0-5 https://cran.r-project.org/web/packages/rscproxy/index.html

rsdmx 0.5-11 https://cran.r-project.org/web/packages/rsdmx/index.html

RSGHB 1.1.2 https://cran.r-project.org/web/packages/RSGHB/index.html

RSiteCatalyst 1.4.14 https://cran.r-

project.org/web/packages/RSiteCatalyst/index.html

RSNNS 0.4-10 https://cran.r-project.org/web/packages/RSNNS/index.html

Rsolnp 1.16 https://cran.r-project.org/web/packages/Rsolnp/index.html

RSpectra 0.12-0 https://cran.r-project.org/web/packages/RSpectra/index.html

RSQLite 2.1.0 https://cran.r-project.org/web/packages/RSQLite/index.html

rstan 2.17.3 https://cran.r-project.org/web/packages/rstan/index.html

rstudioapi 0.7 https://cran.r-project.org/web/packages/rstudioapi/index.html

rsvg 1.1 https://cran.r-project.org/web/packages/rsvg/index.html

RTextTools 1.4.2 https://cran.r-project.org/web/packages/RTextTools/index.html

Rttf2pt1 1.3.6 https://cran.r-project.org/web/packages/Rttf2pt1/index.html

RUnit 0.4.31 https://cran.r-project.org/web/packages/RUnit/index.html

runjags 2.0.4-2 https://cran.r-project.org/web/packages/runjags/index.html

Runuran 0.24 https://cran.r-project.org/web/packages/Runuran/index.html

rvcheck 0.0.9 https://cran.r-project.org/web/packages/rvcheck/index.html

rvest 0.3.2 https://cran.r-project.org/web/packages/rvest/index.html

rworldmap 1.3-6 https://cran.r-project.org/web/packages/rworldmap/index.html

rworldxtra 1.01 https://cran.r-project.org/web/packages/rworldxtra/index.html

SampleSizeMeans 1.1 https://cran.r-

project.org/web/packages/SampleSizeMeans/index.html

SampleSizeProportions 1.0 https://cran.r-

project.org/web/packages/SampleSizeProportions/index.html

sandwich 2.4-0 https://cran.r-project.org/web/packages/sandwich/index.html

Package Version Link

sas7bdat 0.5 https://cran.r-project.org/web/packages/sas7bdat/index.html

satellite 1.0.1 https://cran.r-project.org/web/packages/satellite/index.html

sbgcop 0.975 https://cran.r-project.org/web/packages/sbgcop/index.html

scales 1.0.0 https://cran.r-project.org/web/packages/scales/index.html

scatterplot3d 0.3-41 https://cran.r-

project.org/web/packages/scatterplot3d/index.html

sciplot 1.1-1 https://cran.r-project.org/web/packages/sciplot/index.html

segmented 0.5-3.0 https://cran.r-project.org/web/packages/segmented/index.html

selectr 0.4-0 https://cran.r-project.org/web/packages/selectr/index.html

sem 3.1-9 https://cran.r-project.org/web/packages/sem/index.html

sentimentr 2.7.1 https://cran.r-project.org/web/packages/sentimentr/index.html

seqinr 3.6-1 https://cran.r-project.org/web/packages/seqinr/index.html

seriation 1.2-3 https://cran.r-project.org/web/packages/seriation/index.html

setRNG 2013.9-1 https://cran.r-project.org/web/packages/setRNG/index.html

sf 0.7-4 https://cran.r-project.org/web/packages/sf/index.html

sfsmisc 1.1-2 https://cran.r-project.org/web/packages/sfsmisc/index.html

sgeostat 1.0-27 https://cran.r-project.org/web/packages/sgeostat/index.html

shape 1.4.4 https://cran.r-project.org/web/packages/shape/index.html

shapefiles 0.7 https://cran.r-project.org/web/packages/shapefiles/index.html

shiny 1.0.5 https://cran.r-project.org/web/packages/shiny/index.html

shinyBS 0.61 https://cran.r-project.org/web/packages/shinyBS/index.html

shinycssloaders 0.2.0 https://cran.r-

project.org/web/packages/shinycssloaders/index.html

shinyjs 1.0 https://cran.r-project.org/web/packages/shinyjs/index.html

shinyTime 0.2.1 https://cran.r-project.org/web/packages/shinyTime/index.html

showtext 0.5-1 https://cran.r-project.org/web/packages/showtext/index.html

showtextdb 2.0 https://cran.r-

Package Version Link

project.org/web/packages/showtextdb/index.html

SIS 0.8-6 https://cran.r-project.org/web/packages/SIS/index.html

SixSigma 0.9-51 https://cran.r-project.org/web/packages/SixSigma/index.html

sjlabelled 1.0.8 https://cran.r-project.org/web/packages/sjlabelled/index.html

sjmisc 2.7.1 https://cran.r-project.org/web/packages/sjmisc/index.html

sjPlot 2.4.1 https://cran.r-project.org/web/packages/sjPlot/index.html

sjstats 0.14.2-3 https://cran.r-project.org/web/packages/sjstats/index.html

skmeans 0.2-11 https://cran.r-project.org/web/packages/skmeans/index.html

slam 0.1-42 https://cran.r-project.org/web/packages/slam/index.html

sm 2.2-5.4 https://cran.r-project.org/web/packages/sm/index.html

smooth 2.4.1 https://cran.r-project.org/web/packages/smooth/index.html

smoothSurv 2.0 https://cran.r-

project.org/web/packages/smoothSurv/index.html

sna 2.4 https://cran.r-project.org/web/packages/sna/index.html

snakecase 0.9.1 https://cran.r-project.org/web/packages/snakecase/index.html

snow 0.4-2 https://cran.r-project.org/web/packages/snow/index.html

SnowballC 0.5.1 https://cran.r-project.org/web/packages/SnowballC/index.html

snowFT 1.6-0 https://cran.r-project.org/web/packages/snowFT/index.html

sodium 1.1 https://cran.r-project.org/web/packages/sodium/index.html

sourcetools 0.1.6 https://cran.r-project.org/web/packages/sourcetools/index.html

sp 1.2-7 https://cran.r-project.org/web/packages/sp/index.html

spacetime 1.2-1 https://cran.r-project.org/web/packages/spacetime/index.html

spacyr 0.9.6 https://cran.r-project.org/web/packages/spacyr/index.html

spam 2.1-3 https://cran.r-project.org/web/packages/spam/index.html

SparseM 1.77 https://cran.r-project.org/web/packages/SparseM/index.html

sparsepp 0.2.0 https://cran.r-project.org/web/packages/sparsepp/index.html

spatial 7.3-11 https://cran.r-project.org/web/packages/spatial/index.html

Package Version Link

spatstat 1.55-0 https://cran.r-project.org/web/packages/spatstat/index.html

spatstat.data 1.2-0 https://cran.r-

project.org/web/packages/spatstat.data/index.html

spatstat.utils 1.8-0 https://cran.r-

project.org/web/packages/spatstat.utils/index.html

spBayes 0.4-1 https://cran.r-project.org/web/packages/spBayes/index.html

spData 0.2.8.3 https://cran.r-project.org/web/packages/spData/index.html

spdep 0.7-4 https://cran.r-project.org/web/packages/spdep/index.html

spikeslab 1.1.5 https://cran.r-project.org/web/packages/spikeslab/index.html

splancs 2.01-40 https://cran.r-project.org/web/packages/splancs/index.html

splines 3.4.4 https://cran.r-project.org/web/packages/splines/index.html

spls 2.2-2 https://cran.r-project.org/web/packages/spls/index.html

splus2R 1.2-2 https://cran.r-project.org/web/packages/splus2R/index.html

spTimer 3.0-1 https://cran.r-project.org/web/packages/spTimer/index.html

sqldf 0.4-11 https://cran.r-project.org/web/packages/sqldf/index.html

SQUAREM 2017.10-1 https://cran.r-project.org/web/packages/SQUAREM/index.html

sROC 0.1-2 https://cran.r-project.org/web/packages/sROC/index.html

stabledist 0.7-1 https://cran.r-project.org/web/packages/stabledist/index.html

stabs 0.6-3 https://cran.r-project.org/web/packages/stabs/index.html

StanHeaders 2.17.2 https://cran.r-

project.org/web/packages/StanHeaders/index.html

statmod 1.4.30 https://cran.r-project.org/web/packages/statmod/index.html

statnet.common 4.0.0 https://cran.r-

project.org/web/packages/statnet.common/index.html

stats 3.4.4 NA

stats4 3.4.4 NA

stepPlr 0.93 https://cran.r-project.org/web/packages/stepPlr/index.html

stinepack 1.4 https://cran.r-project.org/web/packages/stinepack/index.html

Package Version Link

stochvol 1.3.3 https://cran.r-project.org/web/packages/stochvol/index.html

stopwords 0.9.0 https://cran.r-project.org/web/packages/stopwords/index.html

stringdist 0.9.4.7 https://cran.r-project.org/web/packages/stringdist/index.html

stringi 1.1.7 https://cran.r-project.org/web/packages/stringi/index.html

stringr 1.3.0 https://cran.r-project.org/web/packages/stringr/index.html

strucchange 1.5-1 https://cran.r-

project.org/web/packages/strucchange/index.html

stsm 1.9 https://cran.r-project.org/web/packages/stsm/index.html

stsm.class 1.3 https://cran.r-project.org/web/packages/stsm.class/index.html

sugrrants 0.2.4 https://cran.r-project.org/web/packages/sugrrants/index.html

sunburstR 2.0.0 https://cran.r-project.org/web/packages/sunburstR/index.html

SuppDists 1.1-9.4 https://cran.r-project.org/web/packages/SuppDists/index.html

survey 3.33-2 https://cran.r-project.org/web/packages/survey/index.html

survival 2.41-3 https://cran.r-project.org/web/packages/survival/index.html

survminer 0.4.6 https://cran.r-project.org/web/packages/survminer/index.html

survMisc 0.5.4 https://cran.r-project.org/web/packages/survMisc/index.html

svglite 1.2.1 https://cran.r-project.org/web/packages/svglite/index.html

svmpath 0.955 https://cran.r-project.org/web/packages/svmpath/index.html

svUnit 0.7-12 https://cran.r-project.org/web/packages/svUnit/index.html

sweep 0.2.1 https://cran.r-project.org/web/packages/sweep/index.html

sysfonts 0.7.2 https://cran.r-project.org/web/packages/sysfonts/index.html

systemfit 1.1-20 https://cran.r-project.org/web/packages/systemfit/index.html

syuzhet 1.0.4 https://cran.r-project.org/web/packages/syuzhet/index.html

tau 0.0-20 https://cran.r-project.org/web/packages/tau/index.html

tcltk 3.4.4 https://cran.r-project.org/web/packages/tcltk/index.html

tcltk2 1.2-11 https://cran.r-project.org/web/packages/tcltk2/index.html

Package Version Link

TeachingDemos 2.10 https://cran.r-

project.org/web/packages/TeachingDemos/index.html

tensor 1.5 https://cran.r-project.org/web/packages/tensor/index.html

tensorA 0.36 https://cran.r-project.org/web/packages/tensorA/index.html

testthat 2.0.0 https://cran.r-project.org/web/packages/testthat/index.html

text2vec 0.5.1 https://cran.r-project.org/web/packages/text2vec/index.html

textcat 1.0-5 https://cran.r-project.org/web/packages/textcat/index.html

textclean 0.9.3 https://cran.r-project.org/web/packages/textclean/index.html

textir 2.0-5 https://cran.r-project.org/web/packages/textir/index.html

textmineR 2.1.1 https://cran.r-project.org/web/packages/textmineR/index.html

textshape 1.6.0 https://cran.r-project.org/web/packages/textshape/index.html

tfplot 2015.12-1 https://cran.r-project.org/web/packages/tfplot/index.html

tframe 2015.12-1 https://cran.r-project.org/web/packages/tframe/index.html

tgp 2.4-14 https://cran.r-project.org/web/packages/tgp/index.html

TH.data 1.0-8 https://cran.r-project.org/web/packages/TH.data/index.html

threejs 0.3.1 https://cran.r-project.org/web/packages/threejs/index.html

tibble 2.1.1 https://cran.r-project.org/web/packages/tibble/index.html

tibbletime 0.1.1 https://cran.r-project.org/web/packages/tibbletime/index.html

tidycensus 0.4.1 https://cran.r-project.org/web/packages/tidycensus/index.html

tidyr 1.0.0 https://cran.r-project.org/web/packages/tidyr/index.html

tidyselect 0.2.5 https://cran.r-project.org/web/packages/tidyselect/index.html

tidytext 0.1.8 https://cran.r-project.org/web/packages/tidytext/index.html

tidyverse 1.2.1 https://cran.r-project.org/web/packages/tidyverse/index.html

tiff 0.1-5 https://cran.r-project.org/web/packages/tiff/index.html

tigris 0.6.2 https://cran.r-project.org/web/packages/tigris/index.html

timeDate 3043.102 https://cran.r-project.org/web/packages/timeDate/index.html

timelineS 0.1.1 https://cran.r-project.org/web/packages/timelineS/index.html

Package Version Link

timeSeries 3042.102 https://cran.r-project.org/web/packages/timeSeries/index.html

timetk 0.1.0 https://cran.r-project.org/web/packages/timetk/index.html

timevis 0.5 https://cran.r-project.org/web/packages/timevis/index.html

tm 0.7-3 https://cran.r-project.org/web/packages/tm/index.html

tmap 1.11-1 https://cran.r-project.org/web/packages/tmap/index.html

tmaptools 1.2-3 https://cran.r-project.org/web/packages/tmaptools/index.html

TMB 1.7.13 https://cran.r-project.org/web/packages/TMB/index.html

tokenizers 0.2.1 https://cran.r-project.org/web/packages/tokenizers/index.html

tools 3.4.4 NA

topicmodels 0.2-7 https://cran.r-

project.org/web/packages/topicmodels/index.html

TraMineR 2.0-8 https://cran.r-project.org/web/packages/TraMineR/index.html

translations 3.4.4 NA

tree 1.0-39 https://cran.r-project.org/web/packages/tree/index.html

treemap 2.4-2 https://cran.r-project.org/web/packages/treemap/index.html

trelliscopejs 0.1.18 https://cran.r-

project.org/web/packages/trelliscopejs/index.html

trimcluster 0.1-2 https://cran.r-project.org/web/packages/trimcluster/index.html

truncnorm 1.0-8 https://cran.r-project.org/web/packages/truncnorm/index.html

TSA 1.01 https://cran.r-project.org/web/packages/TSA/index.html

tseries 0.10-43 https://cran.r-project.org/web/packages/tseries/index.html

tsfa 2014.10-1 https://cran.r-project.org/web/packages/tsfa/index.html

tsibble 0.8.5 https://cran.r-project.org/web/packages/tsibble/index.html

tsintermittent 1.9 https://cran.r-

project.org/web/packages/tsintermittent/index.html

tsoutliers 0.6-6 https://cran.r-project.org/web/packages/tsoutliers/index.html

TSP 1.1-5 https://cran.r-project.org/web/packages/TSP/index.html

Package Version Link

TSstudio 0.1.5 https://cran.r-project.org/web/packages/TSstudio/index.html

TTR 0.23-3 https://cran.r-project.org/web/packages/TTR/index.html

tweedie 2.3.2 https://cran.r-project.org/web/packages/tweedie/index.html

tweenr 1.0.1 https://cran.r-project.org/web/packages/tweenr/index.html

twitteR 1.1.9 https://cran.r-project.org/web/packages/twitteR/index.html

udpipe 0.5 https://cran.r-project.org/web/packages/udpipe/index.html

udunits2 0.13 https://cran.r-project.org/web/packages/udunits2/index.html

units 0.6-2 https://cran.r-project.org/web/packages/units/index.html

UpSetR 1.3.3 https://cran.r-project.org/web/packages/UpSetR/index.html

urca 1.3-0 https://cran.r-project.org/web/packages/urca/index.html

useful 1.2.3 https://cran.r-project.org/web/packages/useful/index.html

UsingR 2.0-5 https://cran.r-project.org/web/packages/UsingR/index.html

usmap 0.2.1 https://cran.r-project.org/web/packages/usmap/index.html

utf8 1.1.3 https://cran.r-project.org/web/packages/utf8/index.html

utils 3.4.4 NA

uuid 0.1-2 https://cran.r-project.org/web/packages/uuid/index.html

V8 2.2 https://cran.r-project.org/web/packages/V8/index.html

vars 1.5-2 https://cran.r-project.org/web/packages/vars/index.html

vcd 1.4-4 https://cran.r-project.org/web/packages/vcd/index.html

vctrs 0.2.0 https://cran.r-project.org/web/packages/vctrs/index.html

vdiffr 0.2.2 https://cran.r-project.org/web/packages/vdiffr/index.html

vegan 2.4-6 https://cran.r-project.org/web/packages/vegan/index.html

VennDiagram 1.6.20 https://cran.r-

project.org/web/packages/VennDiagram/index.html

VGAM 1.0-5 https://cran.r-project.org/web/packages/VGAM/index.html

VIF 1.0 https://cran.r-project.org/web/packages/VIF/index.html

VIM 4.7.0 https://cran.r-project.org/web/packages/VIM/index.html

Package Version Link

vioplot 0.2 https://cran.r-project.org/web/packages/vioplot/index.html

viridis 0.5.1 https://cran.r-project.org/web/packages/viridis/index.html

viridisLite 0.3.0 https://cran.r-project.org/web/packages/viridisLite/index.html

visNetwork 2.0.3 https://cran.r-project.org/web/packages/visNetwork/index.html

vistime 0.4.0 https://cran.r-project.org/web/packages/vistime/index.html

waterfalls 0.1.2 https://cran.r-project.org/web/packages/waterfalls/index.html

wavethresh 4.6.8 https://cran.r-project.org/web/packages/wavethresh/index.html

webshot 0.5.0 https://cran.r-project.org/web/packages/webshot/index.html

webutils 0.6 https://cran.r-project.org/web/packages/webutils/index.html

weco 1.1 https://cran.r-project.org/web/packages/weco/index.html

WeibullR 1.0.10 https://cran.r-project.org/web/packages/WeibullR/index.html

weights 0.85 https://cran.r-project.org/web/packages/weights/index.html

whisker 0.3-2 https://cran.r-project.org/web/packages/whisker/index.html

withr 2.1.2 https://cran.r-project.org/web/packages/withr/index.html

wmtsa 2.0-3 https://cran.r-project.org/web/packages/wmtsa/index.html

wordcloud 2.5 https://cran.r-project.org/web/packages/wordcloud/index.html

wordcloud2 0.2.1 https://cran.r-

project.org/web/packages/wordcloud2/index.html

xesreadR 0.2.2 https://cran.r-project.org/web/packages/xesreadR/index.html

xgboost 0.6.4.1 https://cran.r-project.org/web/packages/xgboost/index.html

XML 3.98-1.10 https://cran.r-project.org/web/packages/XML/index.html

xml2 1.2.0 https://cran.r-project.org/web/packages/xml2/index.html

xplorerr 0.1.1 https://cran.r-project.org/web/packages/xplorerr/index.html

xtable 1.8-2 https://cran.r-project.org/web/packages/xtable/index.html

xts 0.10-2 https://cran.r-project.org/web/packages/xts/index.html

yaml 2.1.18 https://cran.r-project.org/web/packages/yaml/index.html

yarrr 0.1.5 https://cran.r-project.org/web/packages/yarrr/index.html

 Package Version Link

YieldCurve 4.1 https://cran.r-project.org/web/packages/YieldCurve/index.html

zeallot 0.1.0 https://cran.r-project.org/web/packages/zeallot/index.html

zic 0.9.1 https://cran.r-project.org/web/packages/zic/index.html

zipfR 0.6-10 https://cran.r-project.org/web/packages/zipfR/index.html

zoo 1.8-1 https://cran.r-project.org/web/packages/zoo/index.html

R scripts that aren't supported in Power BI

The following table shows which packages are not supported in the Power BI service.

Package Request Date Reason

RgoogleMaps 10/05/2016 Networking is blocked

mailR 10/03/2016 Networking is blocked

RevoScaleR 8/30/2016 Ships only with Microsoft R Server

Next steps
For more information about R in Power BI, take a look at the following articles:

Creating R visuals in the Power BI service

Create Power BI visuals using R
Running R scripts in Power BI Desktop
Using R in Power Query Editor
Enter data directly into Power BI
Desktop
Article • 03/20/2023

With Power BI Desktop, you can enter data directly and use that data in your reports
and visualizations. For example, you can copy portions of a workbook or web page, then
paste it into Power BI Desktop.

To enter data directly into Power BI Desktop in the form of a new table, select Enter data
from the Home ribbon.

Power BI Desktop might attempt to make minor transformations on the data, if

appropriate, just like it does when you load data from any source. For example, in the
following case it promoted the first row of data to headers.
If you want to shape the data you entered or pasted, select Edit to open Power Query
Editor. You can shape and transform the data before bringing it into Power BI Desktop.
Select Load to import the data as it appears.

When you select Load, Power BI Desktop creates a new table from your data, and makes
it available in the Fields pane. In the following image, Power BI Desktop shows your new
table, called Table, and the two fields within that table that were created.

And that’s it. It's that easy to enter data into Power BI Desktop.

You're now ready to use the data in Power BI Desktop. You can create visuals, reports, or
interact with any other data you might want to connect with and import, such as Excel
workbooks, databases, or any other data source.

７ Note

To update, add, or delete data within items created by Enter Data, changes must be
made in Power BI Desktop, and published. Data updates cannot be made directly
from the Power BI service.

Next steps
There are all sorts of data you can connect to using Power BI Desktop. For more
information on data sources, check out the following resources:

What is Power BI Desktop?

Data sources in Power BI Desktop
Shape and combine data with Power BI Desktop
Connect to Excel workbooks in Power BI Desktop
Connect to CSV files in Power BI Desktop
Connect to SSAS multidimensional
models in Power BI Desktop
Article • 01/23/2023

With Power BI Desktop, you can access SQL Server Analysis Services (SSAS)
multidimensional models, commonly referred to as SSAS MD.

To connect to an SSAS MD database, select Get data, choose Database > SQL Server
Analysis Services database, and then select Connect:

The Power BI service and Power BI Desktop both support SSAS multidimensional models
in live connection mode. You can publish and upload reports that use SSAS
Multidimensional models in live mode to the Power BI service.

Capabilities and features of SSAS MD

The following sections describe features and capabilities of Power BI and SSAS MD
connections.
Tabular metadata of multidimensional models
The following table shows the correspondence between multidimensional objects and
the tabular metadata that's returned to Power BI Desktop. Power BI queries the model
for tabular metadata. Based on the returned metadata, Power BI Desktop runs
appropriate DAX queries against SSAS when you create a visualization, such as a table,
matrix, chart, or slicer.

BISM-Multidimensional object Tabular Metadata

Cube Model

Cube dimension Table

Dimension attributes (keys), name Columns

Measure group Table

Measure Measure

Measures without associated measure group Within table called Measures

Measure group -> Cube dimension relationship Relationship

Perspective Perspective

KPI KPI

User/parent-child hierarchies Hierarchies

Measures, measure groups, and KPIs

Measure groups in a multidimensional cube are exposed as tables with a sigma (∑)
beside them in the Fields pane. Calculated measures without an associated measure
group are grouped under a special table called Measures in the tabular metadata.

To help simplify complex models in a multidimensional model, you can define a set of
measures or KPIs in a cube to be located within a display folder. Power BI recognizes
display folders in tabular metadata, and it shows measures and KPIs within the display
folders. KPIs in multidimensional databases support Value, Goal, Status Graphic, and
Trend Graphic.

Dimension attribute type

Multidimensional models also support associating dimension attributes with specific
dimension attribute types. For example, a Geography dimension where the City, State-
Province, CountryRegion, and Postal Code dimension attributes have appropriate
geography types associated with them are exposed in the tabular metadata. Power BI
recognizes the metadata, enabling you to create map visualizations. You can recognize
these associations by the map icon next to the element in the Field pane in Power BI.

Power BI can also render images when you provide a field that contains uniform
resource locators (URLs) of the images. You might specify these fields as ImageURL types
in SQL Server Data Tools, or then in Power BI Desktop. Its type information is then
provided to Power BI in the tabular metadata. Power BI can then retrieve those images
from the URL and display them in visuals.

Parent-child hierarchies
Multidimensional models support parent-child hierarchies, which are presented as a
hierarchy in the tabular metadata. Each level of the parent-child hierarchy is exposed as
a hidden column in the tabular metadata. The key attribute of the parent-child
dimension isn't exposed in the tabular metadata.

Dimension calculated members

Multidimensional models support creation of various types of calculated members. The
two most common types of calculated members are:

Calculated members on attribute hierarchies that aren't siblings of All

Calculated members on user hierarchies

Multidimensional models expose calculated members on attribute hierarchies as values

of a column. You have a few other options and constraints if you expose this type of
calculated member:

A dimension attribute can have an optional UnknownMember.

An attribute containing calculated members can't be the key attribute of the

dimension unless it's the only attribute of the dimension.

An attribute containing calculated members can't be a parent-child attribute.

The calculated members of user hierarchies aren't exposed in Power BI. You can instead
connect to a cube that contains calculated members on user hierarchies. However, you
can't see calculated members if they don't meet the constraints that are mentioned in
the previous bulleted list.

Security
Multidimensional models support dimension and cell level security by way of roles.
When you connect to a cube with Power BI, you're authenticated and evaluated for
appropriate permissions. If a user has dimension security applied, the respective
dimension members aren't seen by the user in Power BI. However, when a user has
defined a cell security permission where certain cells are restricted, that user can't
connect to the cube using Power BI.

Considerations and limitations

There are certain limitations to using SSAS MD:

Only enterprise and BI editions of SQL Server 2014 support live connections. For
the standard edition of SQL Server, SQL Server 2016 or later is required for live
connections.

Actions and named sets aren't exposed to Power BI. To create visuals and reports,
you can still connect to cubes that also contain actions or named sets.

When Power BI displays metadata for an SSAS model, occasionally you can't
retrieve data from the model. This scenario can occur if you've installed the 32-bit
version of the Microsoft Online Analytical Processing provider, but not the 64-bit
version. Installing the 64-bit version might resolve the issue.

You can't create report level measures when authoring a report that is connected
live to an SSAS multidimensional model. The only measures that are available are
measures defined in the MD model.

Supported features of SSAS MD in Power BI

Desktop
Consumption of the following elements is supported in this release of SSAS MD. For
more information about these features, see Understanding power view for
multidimensional models.

Default members
Dimension attributes
Dimension attribute types
Dimension calculated members, which:
must be a single real member when the dimension has more than one attribute;
can't be the key attribute of the dimension unless it's the only attribute; and
can't be a parent-child attribute.
 Dimension security
Display folders
Hierarchies
ImageUrls
KPIs
KPI trends
Measures (with or without measure groups)
Measures as variant

Troubleshooting
The following list describes all known issues when connecting to SQL Server Analysis
Services.

Error : Couldn't load model schema. This error usually occurs when the user
connecting to Analysis Services doesn't have access to database/cube.
Connect to webpages from Power BI
Desktop
Article • 03/20/2023

You can connect to a webpage and import its data into Power BI Desktop, to use in your
visuals and in your data models.

In Power BI Desktop, select Get data > Web from the Home ribbon.

A dialog appears, asking for the URL of the webpage from which you want to import
data.
Once you've typed or pasted the URL, select OK.

Power BI Desktop connects to the webpage and then presents the page's available data
in the Navigator window. When you select one of the available data elements, such as
Table 1, the Navigator window displays a preview of that data on the right side of the
window.

You can choose the Transform Data button, which launches Power Query Editor, where
you can shape and transform the data on that webpage before importing it into Power
BI Desktop. Or you can select the Load button, and import all of the data elements you
selected in the left pane.
When you select Load, Power BI Desktop imports the selected items, and makes them
available in the Fields pane, found on the right side of the Reports view in Power BI
Desktop.

That's all there is to connecting to a webpage and bringing its data into Power BI
Desktop.

From there, you can drag those fields onto the Report canvas and create all the
visualizations you want. You can also use the data from that webpage just like you
would any other data. You can shape it, you can create relationships between it and
other data sources in your model, and otherwise do what you like to create the Power BI
report you want.

To see connecting to a webpage in more depth and action, take a look at the Power BI
Desktop Getting Started Guide.

Certificate revocation check

Power BI applies security for web connections to protect your data. In some scenarios,
such as capturing web requests with Fiddler, web connections may not work properly. To
enable such scenarios, you can modify the Check if your certificates have been revoked
option in Power BI Desktop, then restart Power BI Desktop.

To change this option, select File > Options and settings > Options, then select
Security in the left pane.
Next steps
There are all sorts of data you can connect to using Power BI Desktop. For more
information on data sources, check out the following resources:

Data Sources in Power BI Desktop

Shape and Combine Data with Power BI Desktop
Connect to Excel workbooks in Power BI Desktop
Connect to CSV files in Power BI Desktop
Enter data directly into Power BI Desktop
Connect to Snowflake in the Power BI
service
Article • 12/21/2023

Connecting to Snowflake in the Power BI service differs from other connectors in only
one way. Snowflake has a capability for Microsoft Entra ID, an option for SSO (single
sign-on). Parts of the integration require different administrative roles across Snowflake,
Power BI, and Azure. You can choose to enable Microsoft Entra authentication without
using SSO. Basic authentication works similarly to other connectors in the service.

To configure Microsoft Entra integration and optionally enable SSO:

If you're the Snowflake admin, see Power BI SSO to Snowflake in the Snowflake
documentation.
If you're a Power BI admin, go to the Admin portal section to enable SSO.
If you're a Power BI semantic model creator, go to the Configure a semantic model
with Microsoft Entra ID section to enable SSO.

Power BI service configuration

Admin portal
To enable SSO, a global admin has to turn on the setting in the Power BI Admin portal.
This setting approves sending Microsoft Entra authentication tokens to Snowflake from
within the Power BI service. This setting is set at an organizational level. Follow these
steps to enable SSO:

1. Sign in to Power BI using global admin credentials.

2. Select Settings from the page header menu, then select Admin portal.

3. Select Tenant settings, then scroll to locate Integration settings.

 4. Expand Snowflake SSO, toggle the setting to Enabled, then select Apply.

This step is required to consent to sending your Microsoft Entra token to the Snowflake
servers. After you enable the setting, it can take up to an hour for it to take effect.

After SSO is enabled, you can use reports with SSO.

Configure a semantic model with Microsoft Entra ID

After a report that's based on the Snowflake connector is published to the Power BI
service, the semantic model creator has to update settings for the appropriate
workspace so it can use SSO.

For more information including steps for using Microsoft Entra ID, SSO, and Snowflake,
see Data gateway support for single sign-on with Microsoft Entra ID .

For information about how you can use the on-premises data gateway, see What is an
on-premises data gateway?

If you aren't using the gateway, you're all set. When you have Snowflake credentials
configured on your on-premises data gateway, but you're only using that data source in
your model, switch the Semantic model settings to off on the gateway for that data
model.
To turn on SSO for a semantic model:

1. Sign in to Power BI using semantic model creator credentials.

2. Select the appropriate workspace, then choose Settings from the more options
menu that's located next to the semantic model name.

3. Select Data source credentials and sign in. The semantic model can be signed into
Snowflake with Basic or OAuth2 (Microsoft Entra ID) credentials. By using Microsoft
Entra ID, you can enable SSO in the next step.

4. Select the option End users use their own OAuth2 credentials when accessing
this data source via DirectQuery. This setting will enable Microsoft Entra SSO. The
Microsoft Entra credentials are sent for SSO.
After these steps are done, users should automatically use their Microsoft Entra
authentication to connect to data from that Snowflake semantic model.

If you choose not to enable SSO, then users refreshing the report will use the credentials
of the user who signed in, like most other Power BI reports.

Troubleshooting
If you run into any issues with the integration, see the Snowflake troubleshooting
guide .

Next steps
Data sources for the Power BI service
Connect to semantic models in the Power BI service from Power BI desktop
Connect to Snowflake in Power BI Desktop
Create visuals and reports with the
Microsoft Cost Management connector
in Power BI Desktop
Article • 06/03/2024

You can use the Microsoft Cost Management connector for Power BI Desktop to make
powerful, customized visualizations and reports that help you better understand your
Azure spend.

The Microsoft Cost Management connector currently supports customers with:

A direct Microsoft Customer Agreement

An Enterprise Agreement (EA)
A Microsoft Partner Agreement

If you have an unsupported agreement, you can use Exports to save the cost data to a
share and then connect to it using Power BI. For more information, see Tutorial - Create
and manage exported data from Microsoft Cost Management.

The Microsoft Cost Management connector uses OAuth 2.0 for authentication with
Azure and identifies users who are going to use the connector. Tokens generated in this
process are valid for a specific period. Power BI preserves the token for the next sign-in.
OAuth 2.0, is a standard for the process that goes on behind the scenes to ensure the
secure handling of these permissions. To connect, you must use an Enterprise
Administrator account for Enterprise Agreements, or have appropriate permissions at
the billing account or billing profile levels for Microsoft Customer Agreements.

Connect using Microsoft Cost Management

To use the Microsoft Cost Management connector in Power BI Desktop, take the
following steps:

1. In the Home ribbon, select Get Data.

2. Select Azure from the list of data categories.

3. Select Microsoft Cost Management.

 4. In the dialog that appears, for the Choose Scope drop down, use Manually Input
Scope for Microsoft Customer Agreements, or use Enrollment Number for
Enterprise Agreements (EA).

Connect to a Microsoft Customer Agreement

account
This section describes the steps necessary to connect to a Microsoft Customer
Agreement account.

Connect to a billing account

To connect to a billing account, you need to retrieve your Billing account ID from the
Azure portal:
1. In the Azure portal , navigate to Cost Management + Billing.

2. Select your Billing profile.

3. Under Settings in the menu, select Properties in the sidebar.

4. Under Billing profile, copy the ID.

5. For Choose Scope, select Manually Input Scope and input the connection string as
shown in the following example, replacing {billingAccountId} with the data copied
from the previous steps.
/providers/Microsoft.Billing/billingAccounts/{billingAccountId}

Alternatively, for Choose Scope, select Enrollment Number and input the Billing
Account ID string as copied from the previous steps.

6. Enter the number of months and select OK.

 Alternatively, if you want to download less than a month's worth of data you can
set Number of months to zero, then specify a date range using Start Date and End
Date values that equate to less than 31 days.

7. When prompted, sign in with your Azure user account and password. You must
have access to the Billing account scope to successfully access the billing data.

Connect to a billing profile

To connect to a billing profile, you must retrieve your Billing profile ID and Billing
account ID from the Azure portal:

1. In the Azure portal , navigate to Cost Management + Billing.

2. Select your Billing profile.

3. Under Settings in the menu, select Properties in the sidebar.

4. Under Billing profile, copy the ID.

5. Under Billing account, copy the ID.

 6. For Choose Scope, select Manually Input Scope and input the connection string as
shown in the following example, replacing {billingAccountId} and {billingProfileId}
with the data copied from the previous steps.

/providers/Microsoft.Billing/billingAccounts/{billingAccountId}/billingProfile

s/{billingProfileId}

7. Enter the number of months and select OK.

8. When prompted, sign in with your Azure user account and password. You must
have access to the Billing profile to successfully access the billing profile data.

Connect to an Enterprise Agreement account

To connect with an Enterprise Agreement (EA) account, you can get your enrollment ID
from the Azure portal:

1. In the Azure portal , navigate to Cost Management + Billing.

2. Select your billing account.

3. From the Overview blade, copy the Billing account ID.

4. For Choose Scope, select Enrollment Number.

5. In Scope Identifier paste the billing account ID copied in the previous step.

6. Enter the number of months and then select OK.

7. When prompted, sign in with your Azure user account and password. You must use
an Enterprise Administrator account for Enterprise Agreements.

Data available through the connector

Once you successfully authenticate, a Navigator window appears with the following
available data tables:

ﾉ Expand table

Table Account Supported Scopes Description

Type

Balance summary EA only EA Enrollment Summary of the balance for the current
billing month for Enterprise
Agreements (EA).

Billing events MCA Billing Profile Event log of new invoices, credit
only purchases, etc. Microsoft Customer
Agreement only.
Table Account Supported Scopes Description
Type

Budgets EA, MCA EA Enrollment, MCA Budget details to view actual costs or
Billing Account, usage against existing budget targets.
MCA Billing Profile

Charges MCA MCA Billing Profile A month-level summary of Azure

only usage, Marketplace charges, and
charges billed separately. Microsoft
Customer Agreement only.

Credit lots MCA MCA Billing Profile Azure credit lot purchase details for the
only provided billing profile. Microsoft
Customer Agreement only.

Pricesheets EA, MCA EA Enrollment, MCA Applicable meter rates for the provided
Billing Profile billing profile or EA enrollment.

RI charges EA, MCA EA Enrollment, MCA Charges associated to your Reserved

Billing Profile Instances over the last 24 months. This
table is in the process of being
deprecated, use RI transactions instead

RI EA, MCA EA Enrollment, MCA Reserved Instance purchase

recommendations Billing Profile recommendations based on all your
(shared) subscription usage trends for the last
30 days.

RI EA, MCA EA Enrollment, MCA Reserved Instance purchase

recommendations Billing Profile recommendations based on your single
(single) subscription usage trends for the last
30 days.

RI transactions EA, MCA EA Enrollment, MCA List of transactions for reserved

Billing Profile instances on billing account scope.

RI usage details EA, MCA EA Enrollment, MCA Consumption details for your existing
Billing Profile Reserved Instances over the last month.

RI usage summary EA, MCA EA Enrollment, MCA Daily Azure reservation usage
Billing Profile percentage.

Usage details EA, MCA EA Enrollment, MCA A breakdown of consumed quantities

Billing Account,MCA and estimated charges for the given
Billing Profile billing profile on EA enrollment.

Usage details EA, MCA EA Enrollment, MCA A breakdown of consumed quantities

amortized Billing Account,MCA and estimated amortized charges for
Billing Profile the given billing profile on EA
enrollment.
You can select a table to see a preview dialog. You can select one or more tables by
selecting the boxes beside their name and then select Load.

When you select Load, the data is loaded into Power BI Desktop.

When the data you selected is loaded, the data tables and fields are shown in the Fields
pane.

Considerations and limitations

The following considerations and limitations apply to the Microsoft Cost Management
data connector:

Data row requests exceeding one million rows isn't supported by Power BI. Instead,
you can try using the export feature described in create and manage exported data
in Microsoft Cost Management.

The Microsoft Cost Management data connector doesn't work with Office 365 GCC
customer accounts.

Data refresh: The cost and usage data is typically updated and available in the
Azure portal and supporting APIs within 8 to 24 hours, so we suggest you
 constrain Power BI scheduled refreshes to once or twice a day.

Data source reuse: If you have multiple reports that are pulling the same data, and
don't need more report-specific data transformations, you should reuse the same
data source, which would reduce the amount of time required to pull the Usage
Details data.

For more information on reusing data sources, see the following:

Introduction to semantic models across workspaces
Create reports based on semantic models from different workspaces

You might receive a 400 bad request from the RI usage details when you try to refresh
the data if you've chosen date parameter greater than three months. To mitigate the
error, take the following steps:

1. In Power BI Desktop, select Home > Transform data.

2. In Power Query Editor, select the RI usage details semantic model and select
Advanced Editor.

3. Update the Power Query code as shown in the following paragraphs, which split
the calls into three-month chunks. Make sure you note and retain your enrollment
number, or billing account/billing profile ID.

For EA use the following code update:

let
enrollmentNumber = "<<Enrollment Number>>",
optionalParameters1 = [startBillingDataWindow = "-9",
endBillingDataWindow = "-6"],
source1 = AzureCostManagement.Tables("Enrollment Number",
enrollmentNumber, 5, optionalParameters1),
riusagedetails1 = source1{[Key="riusagedetails"]}[Data],
optionalParameters2 = [startBillingDataWindow = "-6",
endBillingDataWindow = "-3"],
source2 = AzureCostManagement.Tables("Enrollment Number",
enrollmentNumber, 5, optionalParameters2),
riusagedetails2 = source2{[Key="riusagedetails"]}[Data],
riusagedetails = Table.Combine({riusagedetails1, riusagedetails2})
in
riusagedetails

For Microsoft Customer Agreements use the following update:

 let
billingProfileId = "<<Billing Profile Id>>",
optionalParameters1 = [startBillingDataWindow = "-9",
endBillingDataWindow = "-6"],
source1 = AzureCostManagement.Tables("Billing Profile Id",
billingProfileId, 5, optionalParameters1),
riusagedetails1 = source1{[Key="riusagedetails"]}[Data],
optionalParameters2 = [startBillingDataWindow = "-6",
endBillingDataWindow = "-3"],
source2 = AzureCostManagement.Tables("Billing Profile Id",
billingProfileId, 5, optionalParameters2),
riusagedetails2 = source2{[Key="riusagedetails"]}[Data],
riusagedetails = Table.Combine({riusagedetails1, riusagedetails2})
in
riusagedetails

4. Once you've updated the code with the appropriate update from the previous
step, select Done and then select Close & Apply.

You might run into a situation where tags aren't working in the usage details or the tags
column can't be transformed to json. This issue stems from the current UCDD API
returning the tags column by trimming the start and end brackets, which results in
Power BI being unable to transform the column because it returns it as a string. To
mitigate this situation, take the following steps.

1. Navigate to Query Editor.

2. Select the Usage Details table.
3. In the right pane, the Properties pane shows the Applied Steps. You need to add a
custom column to the steps, after the Navigation step.
4. From the menu, select Add column > Add custom column
5. Name the column, for example you could name the column TagsInJson or whatever
you prefer, and then enter the following text in the query:

DAX

```= "{"& [Tags] & "}"

6. Completing the previous steps creates a new column of tags in the json format
7. You can now transfer and expand the column as you need to.

Authentication issues encountered with Microsoft Entra guest accounts: You may have
the appropriate permissions to access the enrollment or billing account, but receive an
authentication error similar to one of the following:
 Access to the resource is forbidden
We couldn’t authenticate with the credentials provided. Please try again.

These errors could be the result of having a user account in a different Microsoft Entra
domain that has been added as a guest user.

For guest accounts: Use the following settings or options as you're prompted with the
authentication dialog when connecting with the Cost Management Power BI connector:

1. Select Sign-in
2. Select the Use another account (bottom of the dialog)
3. Select Sign-in options (bottom of the dialog box)
4. Select Sign into an organization
5. For Domain name, provide the Fully Qualified Domain Name (FQDN) of the
Microsoft Entra domain into which you've been added as a guest.
6. Then, for Pick an account select the user account that you’ve previously
authenticated.

Related content
You can connect to many different data sources using Power BI Desktop. For more
information, see the following articles:

What is Power BI Desktop?

Data Sources in Power BI Desktop
Shape and Combine Data with Power BI Desktop
Connect to Excel workbooks in Power BI Desktop
Enter data directly into Power BI Desktop

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Edit SAP variables in Power BI
Article • 01/30/2024

Report authors who use SAP Business Warehouse or SAP HANA with DirectQuery can
allow end users to edit SAP variables in Power BI Premium and shared workspaces. This
article describes the requirements for editing SAP variables, how to enable this feature,
and how to edit variables in Power BI Desktop and the Power BI service.

Requirements and limitations

The following lists describe the requirements and limitations for editing SAP variables:

Requirements
DirectQuery connection. You must connect to the SAP data source by using
DirectQuery. Import connections aren't supported.

Single sign-on (SSO) set up. You must configure SSO for your gateway for this
feature to work. For more information, see Overview of single sign-on for on-
premises data gateways in Power BI.

Latest gateway version. Make sure to download the latest gateway or update your
existing gateway. For more information, see What is an on-premises data gateway?

Limitations
Multidimensional models only for SAP HANA. For SAP HANA, the SAP edit
variables feature works only with multidimensional models and doesn't work on
relational sources. Ensure you have not selected Treat SAP HANA as a relational
database in Options > Global > DirectQuery > DirectQuery options when editing
SAP HANA variables in Power BI.

No sovereign cloud support. Power Query Online isn't available in sovereign

clouds, so sovereign clouds don't support the edit SAP variables feature.

No mobile support. You can't edit SAP variables in Power BI mobile apps.

Workspace restrictions. Editing SAP variables doesn't work for reports in the
Shared with me tab of My Workspace, or in apps created from V1 workspaces.
Enable editing SAP variables
To enable report users to edit SAP variables:

1. In Power BI Desktop, connect to an SAP HANA or SAP BW data source with a

DirectQuery connection.

2. Go to File > Options and settings > Options, and in the left pane, select
DirectQuery under Current File.

3. Under DirectQuery options in the right pane, select the checkbox next to Allow
end users to change SAP variables for this report.

Edit SAP variables

In Power BI Desktop, you can edit variables by selecting Transform data > Edit variables
in the ribbon. Report creators can add and select variables for the report by using the
following dialog box:

After you publish a report that enables editing SAP variables, the Edit variables link
appears in the Filter pane for the report in the Power BI service. The first time you
publish the report, it might take up to five minutes before the Edit variables link
appears.

７ Note

If the link doesn't appear, manually refresh the semantic model by selecting it from
the list in the Semantic models tab of the workspace, and then selecting the
Refresh icon.

To edit the variables in the Power BI service, report users can:

1. Select Edit variables in the Filter pane for the report.

 2. In the Edit variables dialog box, edit and override the variable values, or select the
Reset button to revert their changes.

Similar to other Power BI persistence behaviors, any changes users make in the Edit
variables dialog box persist only for that user. Selecting Reset to default in the top
menu bar resets the report to its original state, including the variables.
You can change the default variables for reports you own in the Power BI service. If you
own a report that uses SAP HANA or SAP BW and enables editing variables, select Edit
variables to change the variables. When you save the report, the changed variables
become the new default settings for that report. Other users who access the report after
you make the changes see the new settings as the defaults.

Troubleshooting
If you get errors that Power BI can't load data or retrieve data for a visual, or that the
data source connection failed, try the following actions to resolve the error:

In the Power BI service, select Edit variables, set default values for the variables,
and then save the report.

In Power BI Desktop, if you no longer want users to be able to edit variables,

uncheck the option at the report level.

Related content
Use SAP HANA in Power BI Desktop
DirectQuery and SAP Business Warehouse (BW)
DirectQuery and SAP HANA
Use DirectQuery in Power BI
Connect to Analysis Services tabular
data in Power BI Desktop
Article • 01/12/2023

With Power BI Desktop, there are two ways you can connect to and get data from your
SQL Server Analysis Services tabular models:

Explore by using a live connection

Select items and import them into Power BI Desktop

Explore by using a live connection: When you use a live connection, items in your
tabular model or perspective, like tables, columns, and measures, appear in your Power
BI Desktop Fields pane list. You can use Power BI Desktop's advanced visualization and
report tools to explore your tabular model in new, highly interactive ways.

When you connect live, no data from the tabular model is imported into Power BI
Desktop. Each time you interact with a visualization, Power BI Desktop queries the
tabular model and calculates the results that you see. You're always looking at the latest
data that is available in the tabular model, either from the last processing time, or from
DirectQuery tables available in the tabular model.

Keep in mind that tabular models are highly secure. Items that appear in Power BI
Desktop depend on your permissions for the tabular model that you're connected to.

When you've created dynamic reports in Power BI Desktop, you can share them by
publishing to your Power BI site. When you publish a Power BI Desktop file with a live
connection to a tabular model to your Power BI site, an on-premises data gateway must
be installed and configured by an administrator. for more information, see On-premises
data gateway.

Select items and import into Power BI Desktop: When you connect with this option,
you can select items like tables, columns, and measures in your tabular model or
perspective and load them into a Power BI Desktop model. Use Power BI Desktop's
Power Query Editor to further shape what you want and its modeling features to further
model the data. Because no live connection between Power BI Desktop and the tabular
model is maintained, you can then explore your Power BI Desktop model offline or
publish to your Power BI site.

To connect to a tabular model

1. In Power BI Desktop, on the Home ribbon, select Get Data > More > Database.
2. Select SQL Server Analysis Services database, and then select Connect.

3. In the SQL Server Analysis Services database window, enter the Server name,
choose a connection mode, and then select OK.
 4. This step in the Navigator window depends on the connection mode you selected:

If you’re connecting live, select a tabular model or perspective.

If you chose to select items and get data, select a tabular model or
perspective, and then select a particular table or column to load. To shape
your data before loading, select Transform data to open Power Query Editor.
When you’re ready, select Load to import the data into Power BI Desktop.

Frequently asked questions

Question: Do I need an on-premises data gateway?
Answer: It depends. If you use Power BI Desktop to connect live to a tabular model, but
have no intention to publish to your Power BI site, you don't need a gateway. On the
other hand, if you do intend on publishing to your Power BI site, a data gateway is
necessary to ensure secure communication between the Power BI service and your on-
premises Analysis Services server. Be sure to talk to your Analysis Services server
administrator before installing a data gateway.

If you choose to select items and get data, you import tabular model data directly into
your Power BI Desktop file, so no gateway is necessary.

Question: What's the difference between connecting live to a tabular model from the
Power BI service versus connecting live from Power BI Desktop?

Answer: When you connect live to a tabular model from your site in the Power BI service
to an Analysis Services database on-premises in your organization, an on-premises data
gateway is required to secure communications between them. When you connect live to
a tabular model from Power BI Desktop, a gateway isn't required because the Power BI
Desktop and the Analysis Services server you’re connecting to are both running on-
premises in your organization. However, if you publish your Power BI Desktop file to
your Power BI site, a gateway is required.

Question: If I created a live connection, can I connect to another data source in the
same Power BI Desktop file?

Answer: No. You can't explore live data and connect to another type of data source in
the same file. If you’ve already imported data or connected to a different data source in
a Power BI Desktop file, you need to create a new file to explore live.

Question: If I created a live connection, can I edit the model or query in Power BI
Desktop?

Answer: You can create report level measures in the Power BI Desktop, but all other
query and modeling features are disabled when exploring live data.

Question: If I created a live connection, is it secure?

Answer: Yes. Your current Windows credentials are used to connect to the Analysis
Services server. You can't use basic or stored credentials in either the Power BI service or
Power BI Desktop when exploring live.

Question: In Navigator, I see a model and a perspective. What’s the difference?

Answer: A perspective is a particular view of a tabular model. It might include only

particular tables, columns, or measures depending on a unique data analysis need. A
tabular model always contains at least one perspective, which could include everything
in the model. If you’re unsure which perspective you should select, check with your
administrator.

Question: Are there any features of Analysis Services that change the way Power BI
behaves?

Answer: Yes. Depending on the features your tabular model uses, the experience in
Power BI Desktop might change. Some examples include:

You may see measures in the model grouped together at the top of the Fields
pane list rather than in tables alongside columns. Don't worry, you can still use
them as normal, it's just easier to find them this way.

If the tabular model has calculation groups defined, you can use them only with
model measures and not with implicit measures you create by adding numeric
fields to a visual. The model might also have had the DiscourageImplicitMeasures
flag set manually, which has the same effect. For more information, see Calculation
groups in Analysis Services.

To change the server name after initial

connection
After you create a Power BI Desktop file with an explore live connection, there might be
some cases where you want to switch the connection to a different server. For example,
if you created your Power BI Desktop file when connecting to a development server, and
before publishing to the Power BI service, you want to switch the connection to
production server.

To change the server name:

1. Select Transform data > Data source settings from the Home tab.

2. In the SQL Server Analysis Services database window, enter the new Server name,
and then select OK.

Troubleshooting
The following list describes all known issues when connecting to SQL Server Analysis
Services (SSAS) or Azure Analysis Services:

Error: Couldn't load model schema: This error usually occurs when the user
connecting to Analysis Services doesn't have access to the database/model.
Use DirectQuery in Power BI Desktop
Article • 11/10/2023

When you connect to any data source with Power BI Desktop, you can import a copy of
the data. For some data sources, you can also connect directly to the data source
without importing data by using DirectQuery.

To determine whether a data source supports DirectQuery, view the full listing of
available data sources found in the article Connectors in Power Query which also applies
to Power BI, select the article that describes the data source you're interested in from
the list of supported connectors, then see the section in that connector's article titled
Capabilities supported. If DirectQuery isn't listed in that section for the data source's
article, DirectQuery isn't supported for that data connector.

Here are the differences between using import and DirectQuery connectivity modes:

Import: A copy of the data from the selected tables and columns imports into
Power BI Desktop. As you create or interact with visualizations, Power BI Desktop
uses the imported data. To see underlying data changes after the initial import or
the most recent refresh, you must import the full semantic model again to refresh
the data.

DirectQuery: No data imports into Power BI Desktop. For relational sources, you
can select tables and columns to appear in the Power BI Desktop Fields list. For
multidimensional sources like SAP Business Warehouse (SAP BW), the dimensions
and measures of the selected cube appear in the Fields list. As you create or
interact with visualizations, Power BI Desktop queries the underlying data source,
so you're always viewing current data.

With DirectQuery, when you create or interact with a visualization, you must query the
underlying source. The time that's needed to refresh the visualization depends on the
performance of the underlying data source. If the data needed to service the request
was recently requested, Power BI Desktop uses the recent data to reduce the time
required to show the visualization. Selecting Refresh from the Home ribbon refreshes all
visualizations with current data.

Many data modeling and data transformations are available when using DirectQuery,
although with some performance-based limitations. For more information about
DirectQuery benefits, limitations, and recommendations, see DirectQuery in Power BI.

DirectQuery benefits
Some benefits of using DirectQuery include:

DirectQuery lets you build visualizations over very large semantic models, where it
would be unfeasible to import all the data with pre-aggregation.

DirectQuery reports always use current data. Seeing underlying data changes
requires you to refresh the data, and reimporting large semantic models to refresh
data could be unfeasible.

The 1-GB semantic model limitation doesn't apply with DirectQuery.

Connect using DirectQuery

To connect to a data source with DirectQuery:

1. In the Home group of the Power BI Desktop ribbon, select Get data, and then
select a data source that DirectQuery supports, such as SQL Server.

2. In the dialog box for the connection, under Data connectivity mode, select
DirectQuery.

Publish to the Power BI service

You can publish DirectQuery reports to the Power BI service, but you need to take extra
steps for the Power BI service to open the reports.
 To connect the Power BI service to DirectQuery data sources other than Azure SQL
Database, Azure Synapse Analytics (formerly SQL Data Warehouse), Amazon
Redshift, and Snowflake Data Warehouse, install an on-premises data gateway and
register the data source.

If you used DirectQuery with cloud sources like Azure SQL Database, Azure
Synapse, Amazon Redshift, or Snowflake Data Warehouse, you don't need an on-
premises data gateway. You still must provide credentials for the Power BI service
to open the published report. Without credentials, an error occurs when you try to
open a published report or explore a semantic model created with a DirectQuery
connection.

To provide credentials for opening the report and refreshing the data:

1. In the Power BI service, select the gear icon at upper-right and choose Settings.

2. On the Settings page, select the Semantic models tab, and choose the semantic
model that uses DirectQuery.

3. Under Data source connection, provide the credentials to connect to the data
source.

７ Note

If you used DirectQuery with an Azure SQL Database that has a private IP address,
you need to use an on-premises gateway.

Considerations and limitations

Some Power BI Desktop features aren't supported in DirectQuery mode, or they have
limitations. Some capabilities in the Power BI service, such as quick insights, also aren't
available for semantic models that use DirectQuery. When you decide whether to use
DirectQuery, consider these feature limitations. Also consider the following factors:

Performance and load considerations

DirectQuery sends all requests to the source database, so the required refresh time for
visuals depends on how long the underlying source takes to return results. Five seconds
or less is the recommended response time for receiving requested data for visuals.
Refresh times over 30 seconds produce an unacceptably poor experience for users
consuming the report. A query that takes longer than four minutes times out in the
Power BI service, and the user receives an error.

Load on the source database also depends on the number of Power BI users who
consume the published report, especially if the report uses row-level security (RLS). The
refresh of a non-RLS dashboard tile shared by multiple users sends a single query to the
database, but refreshing a dashboard tile that uses RLS requires one query per user. The
increased queries significantly increase load and potentially affect performance.

One-million row limit

DirectQuery defines a one-million row limit for data returned from cloud data sources,
which are any data sources that aren't on-premises. On-premises sources are limited to
a defined payload of about 4 MB per row, depending on proprietary compression
algorithm, or 16 MB for the entire visual. Premium capacities can set different maximum
row limits, as described in the blog post Power BI Premium new capacity settings .

Power BI creates queries that are as efficient as possible, but some generated queries
might retrieve too many rows from the underlying data source. For example, this
situation can occur with a simple chart that includes a high cardinality column with the
aggregation option set to Don't Summarize. The visual must have only columns with a
cardinality below 1 million, or must apply the appropriate filters.

The row limit doesn't apply to aggregations or calculations used to select the semantic
model DirectQuery returns, only to the rows returned. For example, the query that runs
on the data source can aggregate 10 million rows. As long as the data returned to
Power BI is less than 1 million rows, the query can accurately return the results. If the
data is over 1 million rows, Power BI shows an error, except in Premium capacity with
different admin-set limits. The error states: The resultset of a query to external data
source has exceeded the maximum allowed size of '1000000' rows.
Security considerations
By default, all users who consume a published report in the Power BI service connect to
the underlying data source by using the credentials entered after publication. This
situation is the same as for imported data. All users see the same data, regardless of any
security rules that the underlying source defines.

If you need per-user security implemented with DirectQuery sources, either use RLS or
configure Kerberos-constrained authentication against the source. Kerberos isn't
available for all sources. For more information, see Row-level security (RLS) with Power
BI and Configure Kerberos-based SSO from Power BI service to on-premises data
sources.

Other DirectQuery limitations

Some other limitations of using DirectQuery include:

If the Power Query Editor query is overly complex, an error occurs. To fix the error,
you must either delete the problematic step in Power Query Editor, or switch to
import mode. Multidimensional sources like SAP BW can't use the Power Query
Editor.

Automatic date/time hierarchy is unavailable in DirectQuery. DirectQuery mode

doesn't support date column drilldown by year, quarter, month, or day.

For table or matrix visualizations, there's a 125-column limit for results that return
more than 500 rows from DirectQuery sources. These results display a scroll bar in
the table or matrix that lets you fetch more data. In that situation, the maximum
number of columns in the table or matrix is 125. If you must include more than 125
columns in a single table or matrix, consider creating measures that use MIN , MAX ,
FIRST , or LAST , because they don't count against this maximum.

You can't change from import to DirectQuery mode. You can switch from
DirectQuery mode to import mode if you import all the necessary data. It's not
possible to switch back, mostly because of the feature set that DirectQuery doesn't
support. DirectQuery models over multidimensional sources, like SAP BW, can't be
switched from DirectQuery to import mode either, because of the different
treatment of external measures.

Calculated tables and calculated columns that reference a DirectQuery table from a
data source with single sign-on (SSO) authentication aren't supported in the Power
BI service.
Next steps
DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP Business Warehouse (BW)
DirectQuery and SAP HANA
What is an on-premises data gateway?
Using DirectQuery for Power BI semantic models and Azure Analysis Services
(preview)
Connect to SAP Business Warehouse by
using DirectQuery in Power BI
Article • 03/09/2023

You can connect to SAP Business Warehouse (SAP BW) data sources directly using
DirectQuery. Given the OLAP/multidimensional nature of SAP BW, there are many
important differences between DirectQuery over SAP BW versus relational sources like
SQL Server. These differences are summarized as follows:

In DirectQuery over relational sources, there's a set of queries, as defined in the

Get Data or Power Query Editor dialog, that logically defines the data that is
available in the field list. This configuration is not the case when connecting to an
OLAP source such as SAP BW. Instead, when connecting to the SAP server using
Get Data, just the InfoCube or BEx Query is selected. Then all the Key Figures and
dimensions of the selected InfoCube/BEx Query are available in the field list.
Similarly, there's no Power Query Editor when connecting to SAP BW. The data
source settings, for example, server name, can be changed by selecting Transform
data > Data source settings. The settings for any parameters can be changed by
selecting Transform data > Edit parameters.
Given the unique nature of OLAP sources, there are other restrictions for both
modeling and visualizations that apply, in addition to the normal restrictions
imposed for DirectQuery. These restrictions are described later in this article.

In addition, it's extremely important to understand that there are many features of SAP
BW that aren't supported in Power BI, and that because of the nature of the public
interface to SAP BW, there are important cases where the results seen through Power BI
don't match the ones seen when using an SAP tool. These limitations are described later
in this article. These limitations and behavior differences should be carefully reviewed to
ensure that the results seen through Power BI, as returned by the SAP public interface,
are interpreted correctly.

７ Note

The ability to use DirectQuery over SAP BW was in preview until the March 2018
update to Power BI Desktop. During the preview, feedback and suggested
improvements prompted a change that impacts reports that were created using
that preview version. Now that General Availability (GA) of DirectQuery over SAP
BW has released, you must discard any existing (preview-based) reports using
DirectQuery over SAP BW that were created with the pre-GA version.
 In reports created with the pre-GA version of DirectQuery over SAP BW, errors
occur with those pre-GA reports upon invoking Refresh, as a result of attempting to
refresh the metadata with any changes to the underlying SAP BW cube. Please re-
create those reports from a blank report, using the GA version of DirectQuery over
SAP BW.

Additional modeling restrictions

The other primary modeling restrictions when connecting to SAP BW using DirectQuery
in Power BI are:

No support for calculated columns: The ability to create calculated columns is

disabled. This fact also means that grouping and clustering, which create
calculated columns, aren't available.
Additional limitations for measures: There are other limitations imposed on the
DAX expressions that can be used in measures to reflect the level of support
offered by SAP BW.
No support for defining relationships: The relationships are inherent in the
external SAP source. Other relationships can't be defined in the model.
No Data View: The data view normally displays the detail level data in the tables.
Given the nature of OLAP sources like SAP BW, this view isn't available over SAP
BW.
Column and measure details are fixed: The list of columns and measures seen in
the field list are fixed by the underlying source, and can't be modified. For
example, it's not possible to delete a column or change its datatype. It can,
however, be renamed.
Additional limitations in DAX: There are more limitations on the DAX that can be
used in measure definitions to reflect limitations in the source. For example, it's not
possible to use an aggregate function over a table.

Additional visualization restrictions

The other primary restrictions in visualizations when connecting to SAP BW using
DirectQuery in Power BI are:

No aggregation of columns: It's not possible to change the aggregation for a

column on a visual. It's always Do Not Summarize
Measure filtering is disabled: Measure filtering is disabled to reflect the support
offered by SAP BW.
 Multi-select and include/exclude: The ability to multi-select data points on a
visual is disabled if the points represent values from more than one column. For
example, given a bar chart showing Sales by Country/Region, with Category on the
Legend, it wouldn't be possible to select the point for (USA, Bikes) and (France,
Clothes). Similarly, it wouldn't be possible to select the point for (USA, Bikes) and
exclude it from the visual. Both limitations are imposed to reflect the support
offered by SAP BW.

Support for SAP BW features

The following table lists all SAP BW features that aren't fully supported, or behave
differently when using Power BI.

Feature Description

Local calculations Local calculations defined in a BEx Query change the numbers as displayed
through tools like BEx Analyzer. However, they aren't reflected in the
numbers returned from SAP, through the public MDX interface.

As such, the numbers seen in a Power BI visual don't necessarily match

those for a corresponding visual in an SAP tool.

For example, when connecting to a query cube from a BEx query that sets
the aggregation to be Cumulated, or running sum, Power BI would get back
the base numbers, ignoring that setting. An analyst could certainly then
apply a running sum calculation locally in Power BI, but would need to
exercise caution in how the numbers are interpreted if this action isn't done.

Aggregations In some cases, particularly when dealing with multiple currencies, the
aggregate numbers returned by the SAP public interface don't match the
results shown by SAP tools.

As such, the numbers seen in a Power BI visual don't necessarily match

those for a corresponding visual in an SAP tool.

For example, totals over different currencies would show as "*" in BEx
Analyzer, but the total would get returned by the SAP public interface,
without any information that such an aggregate number is meaningless. Thus
the number aggregating, say, $, EUR, and AUD, would get displayed by
Power BI.

Currency Any currency formatting, for example, $2,300 or 4000 AUD, isn't reflected in
formatting Power BI.

Units of measure Units of measure, for example, 230 KG, aren't reflected in Power BI.
Feature Description

Key versus text For an SAP BW characteristic like CostCenter , the field list shows a single
(short, medium, column Cost Center. Using that column displays the default text. By showing
long) hidden fields, it's also possible to see the unique name column that returns
the unique name assigned by SAP BW, and is the basis of uniqueness.

The key and other text fields aren't available.

Multiple In SAP, a characteristic can have multiple hierarchies. Then in tools like BEx
hierarchies of a Analyzer, when a characteristic is included in a query, the user can select the
characteristic hierarchy to use.

In Power BI, the various hierarchies can be seen in the field list as different
hierarchies on the same dimension. However, selecting multiple levels from
two different hierarchies on the same dimension results in empty data being
returned by SAP.

Treatment of
ragged
hierarchies

Scaling In SAP, a key figure can have a scaling factor, for example, 1000, defined as a
factor/reverse formatting option, meaning that all display is scaled by that factor.
sign

It can similarly have a property set that reverses the sign. Use of such a key
figure in Power BI in a visual, or as part of a calculation results in the
unscaled number being used. The sign isn't reversed. The underlying scaling
factor isn't available. In Power BI visuals, the scale units shown on the axis
(K,M,B) can be controlled as part of the visual formatting.

Hierarchies Initially when connecting to SAP BW, the information on the levels of a
where levels hierarchy are retrieved, resulting in a set of fields in the field list. This
appear/disappear information is cached, and if the set of levels changes, then the set of fields
dynamically don't change until Refresh is invoked.
Feature Description

This situation is only possible in Power BI Desktop. Such refresh to reflect

changes to the levels can't be invoked in the Power BI service after publish.

Default filter A BEx query can include default filters, which are applied automatically by
SAP BEx Analyzer. These filters aren't exposed, and hence the equivalent
usage in Power BI doesn't apply the same filters by default.

Hidden Key A BEx query can control visibility of key figures, and those key figures that
figures are hidden don't appear in SAP BEx Analyzer. This fact isn't reflected through
the public API, and hence such hidden key figures still appear in the field list.
However, they can then be hidden within Power BI.

Numeric Any numeric formatting, such as number of decimal positions and decimal
formatting point, isn't automatically reflected in Power BI. However, it's possible to then
control such formatting within Power BI.

Hierarchy SAP BW allows different versions of a hierarchy to be maintained, for

versioning example, the cost center hierarchy in 2007 versus 2008. Only the latest
version is available in Power BI, as information on versions isn't exposed by
the public API.

Time dependent When using Power BI, time dependent hierarchies are evaluated at the
hierarchies current date.

Currency SAP BW supports currency conversion, based on rates held in the cube. Such
conversion capabilities aren't exposed by the public API, and are therefore not available
in Power BI.

Sort Order The sort order, such as by Text or by Key, for a characteristic can be defined in
SAP. This sort order isn't reflected in Power BI. For example, months might
appear as "April", "Aug", and so on.

It's not possible to change this sort order in Power BI.

Technical names In Get Data, the characteristic/measure names (descriptions) and technical
names can both be seen. The field list contains just the
characteristic/measure names (descriptions).

Attributes It's not possible to access the attributes of a characteristic within Power BI.

End user The locale used to connect to SAP BW is set as part of the connection details,
language setting and doesn't reflect the locale of the final report consumer.

Text variables SAP BW allows field names to contain placeholders for variables, for example,
$YEAR$ Actuals , that would then get replaced by the selected value. For
example, the field appears as 2016 Actuals in BEx tools, if the year 2016 were
selected for the variable.
 Feature Description

The column name in Power BI isn't changed depending on the variable value,
and therefore would appear as $YEAR$ Actuals . However, the column name
can then be changed in Power BI.

Customer exit Customer exit variables aren't exposed by the public API, and are therefore
variables not supported by Power BI.

Characteristic Any characteristic structures in the underlying SAP BW source results in an

structures explosion of measures being exposed in Power BI. For example, with two
measures Sales and Costs , and a characteristic structure containing Budget
and Actual, four measures are exposed: Sales.Budget , Sales.Actual ,
Costs.Budget , Costs.Actual .

Next steps
For more information about DirectQuery, check out the following resources:

DirectQuery in Power BI
Data Sources supported by DirectQuery
DirectQuery and SAP HANA
Connect to SAP HANA data sources by
using DirectQuery in Power BI
Article • 07/05/2023

You can connect to SAP HANA data sources directly using DirectQuery. There are two
options when connecting to SAP HANA:

Treat SAP HANA as a multi-dimensional source (default): In this case, the

behavior is similar to when Power BI connects to other multi-dimensional sources
like SAP Business Warehouse, or Analysis Services. When you connect to SAP
HANA using this setting, a single analytic or calculation view is selected and all the
measures, hierarchies and attributes of that view are available in the field list. As
visuals are created, the aggregate data is always retrieved from SAP HANA. This
technique is the recommended approach, and is the default for new DirectQuery
reports over SAP HANA.

Treat SAP HANA as a relational source: In this case, Power BI treats SAP HANA as
a relational source. This approach offers greater flexibility. Care must be taken with
this approach to ensure that measures are aggregated as expected, and to avoid
performance issues.

The connection approach is determined by a global tool option, which is set by selecting
File > Options and settings and then Options > DirectQuery, then selecting the option
Treat SAP HANA as a relational source, as shown in the following image.
The option to treat SAP HANA as a relational source controls the approach used for any
new report using DirectQuery over SAP HANA. It has no effect on any existing SAP
HANA connections in the current report, nor on connections in any other reports that
are opened. So if the option is currently unchecked, then upon adding a new connection
to SAP HANA using Get Data, that connection is made treating SAP HANA as a multi-
dimensional source. However, if a different report is opened that also connects to SAP
HANA, then that report continues to behave according to the option that was set at the
time it was created. This fact means that any reports connecting to SAP HANA that were
created prior to February 2018 continue to treat SAP HANA as a relational source.

The two approaches constitute different behavior, and it's not possible to switch an
existing report from one approach to the other.

Treat SAP HANA as a multi-dimensional source

(default)
All new connections to SAP HANA use this connection method by default, treating SAP
HANA as a multi-dimensional source. In order to treat a connection to SAP HANA as a
relational source, you must select File > Options and settings > Options, then check the
box under Direct Query > Treat SAP HANA as a relational source.

When connecting to SAP HANA as a multi-dimensional source, the following

considerations apply:

In the Get Data Navigator, a single SAP HANA view can be selected. It isn't
possible to select individual measures or attributes. There's no query defined at the
time of connecting, which is different from importing data or when using
DirectQuery while treating SAP HANA as a relational source. This consideration
also means that it's not possible to directly use an SAP HANA SQL query when
selecting this connection method.

All the measures, hierarchies, and attributes of the selected view are displayed in
the field list.

As a measure is used in a visual, SAP HANA is queried to retrieve the measure

value at the level of aggregation necessary for the visual. When dealing with non-
additive measures, such as counters and ratios, all aggregations are performed by
SAP HANA, and no further aggregation is performed by Power BI.

To ensure the correct aggregate values can always be obtained from SAP HANA,
certain restrictions must be imposed. For example, it's not possible to add
calculated columns, or to combine data from multiple SAP HANA views within the
same report.

Treating SAP HANA as a multi-dimensional source doesn't offer the greater flexibility
provided by the alternative relational approach, but it's simpler. The approach also
ensures correct aggregate values when dealing with more complex SAP HANA
measures, and generally results in higher performance.

The Field list includes all measures, attributes, and hierarchies from the SAP HANA view.
Note the following behaviors that apply when using this connection method:

Any attribute that is included in at least one hierarchy is hidden by default.

However, they can be seen if required by selecting View hidden from the context
menu on the field list. From the same context menu they can be made visible, if
necessary.

In SAP HANA, an attribute can be defined to use another attribute as its label. For
example, Product, with values 1 , 2 , 3 , and so on, could use ProductName, with
values Bike , Shirt , Gloves , and so on, as its label. In this case, a single field
 Product is shown in the field list, whose values are the labels Bike , Shirt , Gloves ,
and so on, but which is sorted by, and with uniqueness determined by, the key
values 1 , 2 , 3 . A hidden column Product.Key is also created, allowing access to
the underlying key values if necessary.

Any variables defined in the underlying SAP HANA view are displayed at the time of
connecting, and the necessary values can be entered. Those values can later be changed
by selecting Transform data from the ribbon, and then Edit parameters from the
dropdown menu displayed.

The modeling operations allowed are more restrictive than in the general case when
using DirectQuery, given the need to ensure that correct aggregate data can always be
obtained from SAP HANA. However, it's still possible to make many additions and
changes, including defining measures, renaming and hiding fields, and defining display
formats. All such changes are preserved on refresh, and any non-conflicting changes
made to the SAP HANA view are applied.

Additional modeling restrictions

The other primary modeling restrictions when connecting to SAP HANA using
DirectQuery (treat as multi-dimensional source) are the following restrictions:

No support for calculated columns: The ability to create calculated columns is

disabled. This fact also means that Grouping and Clustering, which create
calculated columns, aren't available.
Additional limitations for measures: There are other limitations imposed on the
DAX expressions that can be used in measures, to reflect the level of support
offered by SAP HANA.
No support for defining relationships: Only a single view can be queried within a
report, and as such, there's no support for defining relationships.
No Data View: The Data View normally displays the detail level data in the tables.
Given the nature of OLAP sources such as SAP HANA, this view isn't available over
SAP HANA.
Column and measure details are fixed: The list of columns and measures seen in
the field list are fixed by the underlying source, and can't be modified. For
example, it's not possible to delete a column, nor change its datatype. It can,
however, be renamed.
Additional limitations in DAX: There are other limitations on the DAX that can be
used in measure definitions, to reflect limitations in the source. For example, it's
not possible to use an aggregate function over a table.
Additional visualization restrictions
There are restrictions in visuals when connecting to SAP HANA using DirectQuery (treat
as multi-dimensional source):

No aggregation of columns: It's not possible to change the aggregation for a

column on a visual, and it's always Do Not Summarize.

Treat SAP HANA as a relational source

When choosing to connect to SAP HANA as a relational source, some extra flexibility
becomes available. For example, you can create calculated columns, include data from
multiple SAP HANA views, and create relationships between the resulting tables.
However, there are differences from the behavior when treating SAP HANA as a
multidimensional source, particularly when the SAP HANA view contains non-additive
measures, for example, distinct counts, or averages, rather than simple sums, and related
to the efficiency of the queries that are run against SAP HANA.

It's useful to start by clarifying the behavior of a relational source such as SQL Server,
when the query defined in Get Data or Power Query Editor performs an aggregation. In
the example that follows, a query defined in Power Query Editor returns the average
price by ProductID.
If the data is being imported into Power BI versus using DirectQuery, the following
situation would result:

The data is imported at the level of aggregation defined by the query created in
Power Query Editor. For example, average price by product. This fact results in a
table with the two columns ProductID and AveragePrice that can be used in visuals.
In a visual, any subsequent aggregation, such as Sum, Average, Min, and others, is
performed over that imported data. For example, including AveragePrice on a
visual uses the Sum aggregate by default, and would return the sum over the
AveragePrice for each ProductID, in this example, 13.67. The same applies to any
alternative aggregate function, such as Min or Average, used on the visual. For
example, Average of AveragePrice returns the average of 6.66, 4 and 3, which
equates to 4.56, and not the average of Price on the six records in the underlying
table, which is 5.17.

If DirectQuery over that same relational source is being used instead of Import, the
same semantics apply and the results would be exactly the same:

Given the same query, logically exactly the same data is presented to the reporting
layer – even though the data isn't actually imported.
 In a visual, any subsequent aggregation, such as Sum, Average, and Min, is again
performed over that logical table from the query. And again, a visual containing
Average of AveragePrice returns the same 4.56.

Consider SAP HANA when the connection is treated as a relational source. Power BI can
work with both Analytic Views and Calculation Views in SAP HANA, both of which can
contain measures. Yet today the approach for SAP HANA follows the same principles as
described previously in this section: the query defined in Get Data or Power Query
Editor determines the data available, and then any subsequent aggregation in a visual is
over that data, and the same applies for both Import and DirectQuery. However, given
the nature of SAP HANA, the query defined in the initial Get Data dialog or Power
Query Editor is always an aggregate query, and generally includes measures where the
actual aggregation that are used is defined by the SAP HANA view.

The equivalent of the previous SQL Server example is that there's an SAP HANA view
containing ID, ProductID, DepotID, and measures including AveragePrice, defined in the
view as Average of Price.

If in the Get Data experience, the selections made were for ProductID and the
AveragePrice measure, then that is defining a query over the view, requesting that
aggregate data. In the earlier example, for simplicity pseudo-SQL is used that doesn’t
match the exact syntax of SAP HANA SQL. Then any further aggregations defined in a
visual are further aggregating the results of such a query. Again, as described previously
for SQL Server, this result applies both for the Import and DirectQuery case. In the
DirectQuery case, the query from Get Data or Power Query Editor are used in a
subselect within a single query sent to SAP HANA, and thus it isn't actually the case that
all the data would be read in, prior to aggregating further.

All of these considerations and behaviors necessitate the following important

considerations when using DirectQuery over SAP HANA:

Attention must be paid to any further aggregation performed in visuals, whenever

the measure in SAP HANA is non-additive, for example, not a simple Sum, Min, or
Max.

In Get Data or Power Query Editor, only the required columns should be included
to retrieve the necessary data, reflecting the fact that the result is a query that
must be a reasonable query that can be sent to SAP HANA. For example, if dozens
of columns were selected, with the thought that they might be needed on
subsequent visuals, then even for DirectQuery a simple visual means the aggregate
query used in the subselect contains those dozens of columns, which generally
perform poorly.
In the following example, selecting five columns (CalendarQuarter, Color, LastName,
ProductLine, SalesOrderNumber) in the Get Data dialog, along with the measure
OrderQuantity, means that later creating a simple visual containing the Min
OrderQuantity results in the following SQL query to SAP HANA. The shaded is the
subselect, containing the query from Get Data / Power Query Editor. If this subselect
gives a high cardinality result, then the resulting SAP HANA performance is likely to be
poor.

Because of this behavior, we recommend the items selected in Get Data or Power Query
Editor be limited to those items that are needed, while still resulting in a reasonable
query for SAP HANA.

Best practices
For both approaches to connecting to SAP HANA, recommendations for using
DirectQuery also apply to SAP HANA, particularly recommendations related to ensuring
good performance. For more information, see using DirectQuery in Power BI.

Considerations and limitations

The following list describes all SAP HANA features that aren't fully supported, or
features that behave differently when using Power BI.

Parent Child Hierarchies: Parent child hierarchies aren't visible in Power BI. This
fact is because Power BI accesses SAP HANA using the SQL interface, and parent
child hierarchies can't be fully accessed by using SQL.
Other hierarchy metadata: The basic structure of hierarchies is displayed in Power
BI, however some hierarchy metadata, such as controlling the behavior of ragged
hierarchies, have no effect. Again, this fact is due to the limitations imposed by the
SQL interface.
Connection using SSL: You can connect using Import and multi-dimensional with
TLS, but can't connect to SAP HANA instances configured to use TLS for the
relational connector.
Support for Attribute views: Power BI can connect to Analytic and Calculation
views, but can't connect directly to Attribute views.
Support for Catalog objects: Power BI can't connect to Catalog objects.
Change to Variables after publish: You can't change the values for any SAP HANA
variables directly in the Power BI service, after the report is published.

Known issues
The following list describes all known issues when connecting to SAP HANA
(DirectQuery) using Power BI.

SAP HANA issue when query for Counters, and other measures: Incorrect data is
returned from SAP HANA if connecting to an Analytical View, and a Counter
measure and some other ratio measure, are included in the same visual. This issue
is covered by SAP Note 2128928 (Unexpected results when query a Calculated
Column and a Counter) . The ratio measure is incorrect in this case.

Multiple Power BI columns from single SAP HANA column: For some calculation
views, where an SAP HANA column is used in more than one hierarchy, SAP HANA
exposes the column as two separate attributes. This approach results in two
columns being created in Power BI. Those columns are hidden by default, however,
and all queries involving the hierarchies, or the columns directly, behave correctly.

Related content
For more information about DirectQuery, check out the following resources:

DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP BW
On-premises data gateway
Apply the Assume Referential Integrity
setting in Power BI Desktop
Article • 01/18/2023

When connecting to a data source using DirectQuery, you can use the Assume
Referential Integrity selection to enable running more efficient queries against your
data source. This feature has a few requirements of the underlying data, and it's only
available when using DirectQuery.

Setting Assume Referential Integrity enables queries on the data source to use INNER
JOIN statements rather than OUTER JOIN, which improves query efficiency.

Requirements for using Assume Referential

Integrity
This setting is an advanced setting, and is only enabled when connecting to data using
DirectQuery. The following requirements are necessary for Assume Referential Integrity
to work properly:

Data in the From column in the relationship is never Null or blank

For each value in the From column, there's a corresponding value in the To column

In this context, the From column is the Many in a One-to-Many relationship, or it's the
column in the first table in a One-to-One relationship.

Example of using Assume Referential Integrity

The following example demonstrates how Assume Referential Integrity behaves when
used in data connections. The example connects to a data source that includes an
Orders table, a Products table, and a Depots table.

In the following image that shows the Orders table and the Products table,
referential integrity exists between Orders[ProductID] and Products[ProductID].
The [ProductID] column in the Orders table is never Null, and every value also
appears in the Products table. As such, Assume Referential Integrity should be set
to get more efficient queries. Using this setting doesn't change the values shown in
visuals.

In the next image, notice that no referential integrity exists between

Orders[DepotID] and Depots[DepotID], because the DepotID is Null for some
Orders. As such, Assume Referential Integrity should not be set.
 Finally, no referential integrity exists between Orders[CustomerID] and
Customers[CustID] in the following tables. The CustomerID contains some values,
in this case, CustX, that don't exist in the Customers table. As such, Assume
Referential Integrity should not be set.

Setting Assume Referential Integrity

To enable this feature, select Assume Referential Integrity as shown in the following
image.
When selected, the setting is validated against the data to ensure there are no Null or
mismatched rows. However, for cases with a very large number of values, the validation
isn't a guarantee that there are no referential integrity issues.

In addition, the validation occurs at the time of editing the relationship, and does not
reflect any subsequent changes to the data.

What happens if you incorrectly set Assume

Referential Integrity?
If you set Assume Referential Integrity when there are referential integrity issues in the
data, that setting doesn't result in errors. However, it does result in apparent
inconsistencies in the data. For example, for the relationship to the Depots table
described here, it would result in the following:

A visual showing the total Order Qty would show a value of 40

A visual showing the total Order Qty by Depot City would show a total value of
only 30, because it wouldn't include Order ID 1, where DepotID is Null.
Next steps
Learn more about DirectQuery.
Get more information about Relationships in Power BI.
Learn more about Relationship View in Power BI Desktop.
Use the SAP Business Warehouse
connector in Power BI Desktop
Article • 03/26/2024

You can use Power BI Desktop to access SAP Business Warehouse (SAP BW) data. The
SAP BW Connector Implementation 2.0 has significant improvements in performance
and capabilities from version 1.0.

For information about how SAP customers can benefit from connecting Power BI to their
SAP BW systems, see the Power BI and SAP BW whitepaper . For details about using
DirectQuery with SAP BW, see DirectQuery and SAP Business Warehouse (BW).

） Important

Version 1.0 of the SAP BW connector is deprecated. New connections use

Implementation 2.0 of the SAP BW connector. All support for version 1.0 will be
removed from the connector in the near future. Use the information in this article
to update existing version 1.0 reports to use Implementation 2.0 of the connector.

Use the SAP BW Connector

Follow these steps to install and connect to data with the SAP BW Connector.

Prerequisite
Implementation 2.0 of the SAP Connector requires the SAP .NET Connector 3.0 or 3.1.
You can download the SAP .NET Connector 3.0 or 3.1 from SAP. Access to the
download requires a valid S-user sign-in.

The .NET Framework connector comes in 32-bit and 64-bit versions. Choose the version
that matches your Power BI Desktop installation version.

When you install, in Optional setup steps, make sure you select Install assemblies to
GAC.
 ７ Note

The first version of the SAP BW Connector required the NetWeaver DLLs. The
current version doesn't require NetWeaver DLLs.

Connect to SAP BW data in Power BI Desktop

To connect to SAP BW data by using the SAP BW Connector, follow these steps:

1. In Power BI Desktop, select Get data.

2. On the Get Data screen, select Database, and then select either SAP Business
Warehouse Application Server or SAP Business Warehouse Message Server.
3. Select Connect.

4. On the next screen, enter server, system, and client information, and whether to
use Import or DirectQuery connectivity method. For detailed instructions, see:

Connect to an SAP BW Application Server from Power Query Desktop

Connect to an SAP BW Message Server from Power Query Desktop

７ Note

You can use the SAP BW Connector to import data from your SAP BW Server
cubes, which is the default, or you can use DirectQuery to connect to the data.
For more information about using the SAP BW Connector with DirectQuery,
see DirectQuery and SAP Business Warehouse (BW).
 You can also select Advanced options, and select a Language code, a custom
MDX statement to run against the specified server, and other options. For more
information, see Use advanced options.

5. Select OK to establish the connection.

6. Provide any necessary authentication data and select Connect. For more
information about authentication, see Authentication with a data source.

7. If you didn't specify a custom MDX statement, the Navigator screen shows a list of
all cubes available on the server. You can drill down and select items from the
available cubes, including dimensions and measures. Power BI shows queries and
cubes that the Open Analysis Interfaces expose.

When you select one or more items from the server, the Navigator shows a
preview of the output table.

The Navigator dialog also provides the following display options:

Only selected items. By default, Navigator displays all items. This option is
useful to verify the final set of items you select. Alternatively, you can select
the column names in the preview area to view the selected items.
Enable data previews. This value is the default, and displays data previews.
Deselect this option to reduce the number of server calls by no longer
 requesting preview data.
Technical names. SAP BW supports user-defined technical names for objects
within a cube. Cube owners can expose these friendly names for cube
objects, instead of exposing only the physical names for the objects.

8. After you select all the objects you want, choose one of the following options:

Load to load the entire set of rows for the output table into the Power BI
Desktop data model. The Report view opens. You can begin visualizing the
data, or make further modifications by using the Data or Model views.
Transform Data to open Power Query Editor with the data. You can specify
more data transformation and filtering steps before you bring the entire set
of rows into the Power BI Desktop data model.

Along with data from SAP BW cubes, you can also import data from a wide range of
other data sources in Power BI Desktop, and combine them into a single report. This
ability presents many interesting scenarios for reporting and analytics on top of SAP BW
data.

New options in SAP BW Implementation 2.0

This section lists some SAP BW Connector Implementation 2.0 features and
improvements. For more information, see Implementation details.

Advanced options
You can set the following options under Advanced options on the SAP BW connection
screen:

Execution mode specifies how the MDX interface executes queries on the server.
The following options are valid:
BasXml
BasXmlGzip
DataStream

The default value is BasXmlGzip. This mode can improve performance for low
latency or high volume queries.

Batch size specifies the maximum number of rows to retrieve at a time when
executing an MDX statement. A small number means more calls to the server while
retrieving a large semantic model. A large value might improve performance, but
could cause memory issues on the SAP BW server. The default value is 50000.

Enable characteristic structures changes the way the Navigator displays

characteristic structures. The default value for this option is false, or unchecked.
This option affects the list of objects available for selection, and isn't supported in
native query mode.

Other improvements
The following list describes other Implementation 2.0 improvements:

Better performance.
Ability to retrieve several million rows of data, and fine-tuning through the batch
size parameter.
Ability to switch execution modes.
Support for compressed mode, especially beneficial for high-latency connections
or large semantic models.
Improved detection of Date variables.
Date (ABAP type DATS ) and Time (ABAP type TIMS ) dimensions exposed as dates

and times, instead of text values. For more information, see Support for typed
dates in SAP BW.
Better exception handling. Errors that occur in BAPI calls are now surfaced.
 Column folding in BasXml and BasXmlGzip modes. For example, if the generated
MDX query retrieves 40 columns but the current selection only needs 10, this
request passes on to the server to retrieve a smaller semantic model.

Update existing Implementation 1.0 reports

You can change existing reports to use Implementation 2.0 only in Import mode.

1. From the existing report in Power BI Desktop, select Transform data in the ribbon,
and then select the SAP Business Warehouse query to update.

2. Right-click the query and select Advanced Editor.

3. In the Advanced Editor, change the SapBusinessWarehouse.Cubes calls as follows:

4. Determine whether the query already contains an option record, such as the
following examples:

If so, add the [Implementation 2.0] option, and remove any ScaleMeasures option:

７ Note

The ScaleMeasures option is deprecated in this implementation. The

connector now always shows unscaled values.

5. If the query doesn't already include an options record, add it. For example, change
the following entry:
 to:

７ Note

Implementation 2.0 of the SAP BW Connector should be compatible with version 1.

However, there might be some differences because of the different SAP BW MDX
execution modes. To resolve any discrepancies, try switching between execution
modes.

Troubleshooting
This section provides some troubleshooting situations and solutions for the SAP BW
connector. For more information, see SAP Business Warehouse connector
troubleshooting.

Numeric data from SAP BW returns misformatted

numeric data
In this issue, SAP BW returns numeric data with decimal points instead of commas. For
example, 1,000,000 returns as 1.000.000.

SAP BW returns decimal data with either a comma or a period as the decimal separator.
To specify which of these characters SAP BW should use for the decimal separator, the
Power BI Desktop driver makes a call to BAPI_USER_GET_DETAIL . This call returns a
structure called DEFAULTS , which has a field called DCPFM that stores Decimal Format
Notation as one of the following values:

' ' (space) = Decimal point is comma: N.NNN,NN

'X' = Decimal point is period: N,NNN.NN

'Y' = Decimal point is N: NNN NNN,NN

With this issue, the call to BAPI_USER_GET_DETAIL fails for a particular user, who gets the
misformatted data, with an error message similar to the following message:

XML

You are not authorized to display users in group TI:

<item>
 <TYPE>E</TYPE>
<ID>01</ID>
<NUMBER>512</NUMBER>
<MESSAGE>You are not authorized to display users in group
TI</MESSAGE>
<LOG_NO/>
<LOG_MSG_NO>000000</LOG_MSG_NO>
<MESSAGE_V1>TI</MESSAGE_V1>
<MESSAGE_V2/>
<MESSAGE_V3/>
<MESSAGE_V4/>
<PARAMETER/>
<ROW>0</ROW>
<FIELD>BNAME</FIELD>
<SYSTEM>CLNTPW1400</SYSTEM>
</item>

To solve this error, the SAP admin must grant the Power BI SAP BW user the right to
execute BAPI_USER_GET_DETAIL . Also, verify that the user's data has the correct DCPFM
value.

Need connectivity for SAP BEx queries

You can do BEx queries in Power BI Desktop by enabling the Release for External Access
property, as shown in the following image:

Navigator doesn't display a data preview

In this issue, Navigator doesn't display a data preview and instead shows an Object
reference not set to an instance of an object error message.

SAP users need access to the following specific BAPI function modules to get metadata
and retrieve data from SAP BW's InfoProviders:

BAPI_MDPROVIDER_GET_CATALOGS
 BAPI_MDPROVIDER_GET_CUBES
BAPI_MDPROVIDER_GET_DIMENSIONS
BAPI_MDPROVIDER_GET_HIERARCHYS
BAPI_MDPROVIDER_GET_LEVELS
BAPI_MDPROVIDER_GET_MEASURES
BAPI_MDPROVIDER_GET_MEMBERS
BAPI_MDPROVIDER_GET_VARIABLES
BAPI_IOBJ_GETDETAIL

To solve this issue, verify that the user has access to the MDPROVIDER modules and
BAPI_IOBJ_GETDETAIL .

Enable tracing
To further troubleshoot these or similar issues, you can enable tracing:

1. In Power BI Desktop, select File > Options and settings > Options.
2. In Options, select Diagnostics, and then select Enable tracing under Diagnostic
Options.
3. Try to get data from SAP BW while tracing is active, and examine the trace file for
more detail.

SAP BW Connection support

The following table describes current Power BI support for SAP BW.

ﾉ Expand table

Product Mode Authentication Connector SNC Library Supported

Power BI Any User / password Application N/A Yes

Desktop Server

Power BI Any Windows Application sapcrypto + Yes

Desktop Server gsskrb5/gx64krb5

Power BI Any Windows via Application sapcrypto + Yes

Desktop impersonation Server gsskrb5/gx64krb5

Power BI Any User / password Message N/A Yes

Desktop Server

Power BI Any Windows Message sapcrypto + Yes

Desktop Server gsskrb5/gx64krb5
Product Mode Authentication Connector SNC Library Supported

Power BI Any Windows via Message sapcrypto + Yes

Desktop impersonation Server gsskrb5/gx64krb5

Power BI Import Same as Power BI

Gateway Desktop

Power BI DirectQuery User / password Application N/A Yes

Gateway Server

Power BI DirectQuery Windows via Application sapcrypto + Yes

Gateway impersonation Server gsskrb5/gx64krb5
(fixed user, no SSO)

Power BI DirectQuery Use SSO via Application sapcrypto + Yes

Gateway Kerberos for Server gsskrb5/gx64krb5
DirectQuery queries
option

Power BI DirectQuery User / password Message N/A Yes

Gateway Server

Power BI DirectQuery Windows via Message sapcrypto + Yes

Gateway impersonation Server gsskrb5/gx64krb5
(fixed user, no SSO)

Power BI DirectQuery Use SSO via Message gsskrb5/gx64krb5 No

Gateway Kerberos for Server
DirectQuery queries
option

Power BI DirectQuery Use SSO via Message sapcrypto Yes

Gateway Kerberos for Server
DirectQuery queries
option

Related content
SAP BW fundamentals
DirectQuery and SAP HANA
DirectQuery and SAP Business Warehouse (BW)
Use DirectQuery in Power BI
Power BI data sources
Power BI and SAP BW whitepaper
Use OneDrive for work or school links in
Power BI Desktop
Article • 11/10/2023

Many people have Excel workbooks stored in OneDrive for work or school that would be
great for use with Power BI Desktop. With Power BI Desktop, you can use online links for
Excel files stored in OneDrive for work or school to create reports and visuals. You can
use a OneDrive for work or school group account or your individual OneDrive for work
or school account.

Getting an online link from OneDrive for work or school requires a few specific steps.
The following sections explain those steps, which let you share the file link among
groups, across different machines, and with your coworkers.

Get a link from Excel

1. Navigate to your OneDrive for work or school location using a browser. Select the
ellipses (...) to open the More menu, then select Details.
 ７ Note

Your browser interface might not look exactly like this image. There are many
ways to select Open in Excel for files in your OneDrive for work or school
browser interface. You can use any option that allows you to open the file in
Excel.

2. In the More details pane that appears, select the copy icon next to Path.
Use the link in Power BI Desktop
In Power BI Desktop, you can use the link that you just copied to the clipboard. Take the
following steps:

1. In Power BI Desktop, select Get data > Web.

2. With the Basic option selected, paste the link into the From Web dialog.

3. If Power BI Desktop prompts you for credentials, choose either Windows for on-
premises SharePoint sites or Organizational Account for Microsoft 365 or
OneDrive for work or school sites.
 A Navigator dialog appears. It allows you to select from the list of tables, sheets,
and ranges found in the Excel workbook. From there, you can use the OneDrive for
work or school file just like any other Excel file. You can create reports and use it in
semantic models like you would with any other data source.

７ Note

To use a OneDrive for work or school file as a data source in the Power BI service,
with Service Refresh enabled for that file, make sure you select OAuth2 as the
Authentication method when you configure refresh settings. Otherwise, you might
encounter an error when you attempt to connect or to refresh, such as, Failed to
update data source credentials. Selecting OAuth2 as the authentication method
avoids that credentials error.
Connect to Project Online data through
Power BI Desktop
Article • 01/19/2023

You can connect to data in Project Online through Power BI Desktop.

Step 1: Download Power BI Desktop

Download Power BI Desktop , then run the installer to get Power BI Desktop on your
computer.

Step 2: Connect to Project Online with OData

1. Open Power BI Desktop.

2. On the Welcome screen, select Get data.

3. Select OData Feed and choose Connect.

4. Enter the address for your OData feed in the URL box, and then select OK.

If the address for your Project Web App site resembles

https://<tenantname>.sharepoint.com/sites/pwa, then the address to enter for your
OData Feed is https://<tenantname>.sharepoint.com/sites/pwa/_api/Projectdata.

5. Power BI Desktop prompts you to authenticate with your work or school account.
Select Organizational account and then enter your credentials.

The account you use to connect to the OData feed must have at least Portfolio Viewer
access to the Project Web App site.
From here, you can choose which tables you would like to connect to and build a query.
Get data from files for Power BI
Article • 11/10/2023

In Power BI, you can connect to or import data and reports from these types of files:

Microsoft Excel .xlsx and .xlsm files

Power BI Desktop .pbix report files
Comma-separated value (CSV) .csv files

What it means to get data from a file

In Power BI, the data you explore comes from a semantic model. To have a semantic
model, you need some data. This article focuses on getting data from files.

To better understand the importance of semantic models and how to get data for them,
consider an automobile. Sitting in your car and looking at the dashboard is like sitting in
front of your computer looking at a dashboard in Power BI. The dashboard shows all the
things your car is doing, like how fast the engine is revving, the temperature, what gear
you’re in, and your speed.

In Power BI, a semantic model is like the engine in your car. The semantic model
provides the data, metrics, and information that's displayed in your Power BI dashboard.
Your engine, or semantic model, needs fuel, and data is the fuel in Power BI. Your car has
a fuel tank that provides gas to the engine. Power BI also needs a fuel tank of data you
can feed your semantic model. That fuel tank can be a Power BI Desktop file, Excel
workbook file, or CSV file.

To take it one step further, a fuel tank in a car has to be filled with gas. The gas for a
Power BI Desktop, Excel, or CSV file is data from a data source that you put into the
Excel, Power BI Desktop, or CSV file. You can manually enter rows of data into an Excel
workbook or CSV file, or you can connect to the external data source to query and load
data into your file. After you have a file that contains some data, you can get the file into
Power BI as a semantic model.

７ Note

When you import Excel data into Power BI, the data must be in a table or data
model.
Where to save your file
Where you save your file makes a difference.

Local. If you save your workbook file to a drive on your computer or another
location in your organization, you can import your file into Power BI. Your file
remains on the source drive. When you import the file, Power BI creates a new
semantic model in your site and loads your data, and in some cases your data
model, into the semantic model. Any reports in your file appear in My workspace
as Reports.

OneDrive for work or school. If you have OneDrive for work or school, sign in with
the same account that you use for Power BI. This method is the most effective way
to keep your work in Excel, Power BI Desktop, or CSV files in sync with your Power
BI semantic model, reports, and dashboards. Both Power BI and OneDrive are in
the cloud, and Power BI connects to your file on OneDrive about once an hour. If
Power BI finds any changes, it automatically updates your Power BI semantic
model, reports, and dashboards.

７ Note

You can't upload files from personal OneDrive accounts, but you can upload
files from your computer.

SharePoint team site. Saving your Power BI Desktop files to a SharePoint team site
is much like saving to OneDrive for work or school. The biggest difference is how
you connect to the file from Power BI. You can specify a URL or connect to the root
folder.

７ Note

You can't update semantic models imported from OneDrive for work or school
from local files. For Power BI to update the semantic model, you must replace the
file in OneDrive for work or school. Alternatively, you can delete the semantic
model and its related items and then import again from a local file.

Next steps
Get data from Excel workbook files
Get data from Power BI Desktop files
Get data from comma-separated value (CSV) files
Get data from Excel workbook files
Article • 11/10/2023

Microsoft Excel is one of the most widely used business applications and one of the
most common data sources for Power BI.

Supported workbooks
Power BI supports importing or connecting to workbooks created in Excel 2007 and
later. Some features that this article describes are available only in later versions of Excel.
Workbooks must be in the .xlsx or .xlsm file type and be smaller than 1 GB.

） Important

The following capabilities are deprecated and will no longer be available starting
September 29th, 2023:

Upload of local workbooks to Power BI workspaces will no longer be allowed.

Configuring scheduling of refresh and refresh now for Excel files that don’t
already have scheduled refresh configured will no longer be allowed.

The following capabilities are deprecated and will no longer be available starting
October 31, 2023:

Scheduled refresh and refresh now for existing Excel files that were previously
configured for scheduled refresh will no longer be allowed.
Local workbooks uploaded to Power BI workspaces will no longer open in
Power BI.

After October 31, 2023:

You can download existing local workbooks from your Power BI workspace.
You can publish your Excel data model as a Power BI semantic model and
schedule refresh.
You can import Excel workbooks from OneDrive and SharePoint Document
libraries to view them in Power BI.

If your organization uses these capabilities, see more details in Migrating your
Excel workbooks.
Workbooks with ranges or tables of data
If your workbook contains simple worksheets with ranges of data, be sure to format
those ranges as tables to get the most out of your data in Power BI. When you create
reports in Power BI, the named tables and columns in the Tables pane make it much
easier to visualize your data.

Workbooks with data models

A workbook can contain a data model that has one or more tables of data loaded into it
via linked tables, Power Query, Get & Transform Data in Excel, or Power Pivot. Power BI
supports all data model properties, like relationships, measures, hierarchies, and key
performance indicators (KPIs).

７ Note

You can't share workbooks that contain data models across Power BI tenants. For
example, a user who signs in to Power BI with a contoso.com account can't share a
workbook containing data models with a user who signs in with a
woodgrovebank.com account.

Workbooks with connections to external data sources

If your Excel workbook connects to an external data source, after your workbook is in
Power BI, you can create reports and dashboards based on data from that connected
source. You can also set up scheduled refresh to automatically connect to the data
source and get updates. You no longer need to refresh manually by using Get Data in
Excel. Visualizations in reports and dashboard tiles that are based on the data source
update automatically. For more information, see Data refresh in Power BI.

Workbooks with PivotTables and charts

Whether and how your PivotTables and charts appear in Power BI depends on where
you save your workbook file, and how you choose to get the file into Power BI. The rest
of this article explains the options.

Data types
Assign data specific data types in Excel to improve your Power BI experience. Power BI
supports these data types:

Whole number
Decimal number
Currency
Date
True/false
Text

Import or upload Excel data

There are two ways to explore Excel data in Power BI: upload and import. When you
upload your workbook, it appears in Power BI just like it would in Excel Online. But you
also have some great features to help you pin elements from your worksheets to your
dashboards. When you import your data, Power BI imports any supported data in tables
and any data model into a new Power BI semantic model.

Upload to Power BI
You can use the Upload button to upload files to the Power BI service. In the workspace
where you want to add the file, select Upload at the top of the page. In the drop-down
list, select:

OneDrive for Business to connect to files that are stored in OneDrive for Business.
SharePoint to connect to files on any SharePoint site that you have access to.
Browse to upload files from your computer.
If you upload a local file, Power BI adds a copy of the file to the workspace. If you use
the OneDrive for Business or SharePoint options, Power BI creates a connection to the
file. As you make changes to the file in SharePoint or OneDrive, Power BI automatically
syncs those changes about once an hour.

When you connect to an Excel file by using OneDrive for Business, you can't edit your
workbook in Power BI. If you need to make changes, you can select Edit and then
choose to edit your workbook in Excel Online or open it in Excel on your computer.
Changes are saved to the workbook on OneDrive.

You should connect to or upload data if you have only data in worksheets, or if you have
ranges, PivotTables, and charts that you want to pin to dashboards.

Local Excel workbooks open in Excel Online within Power BI. Unlike Excel workbooks
stored on OneDrive or SharePoint team sites, you can't edit local Excel files within Power
BI.

If you use Excel 2016 and later, you can also use File > Publish > Upload from Excel. For
more information, see Publish to Power BI from Microsoft Excel.

After your workbook uploads, it appears in the list of content in the workspace:

This upload method is easy to use, and the OneDrive for Business and SharePoint
options use the same file selection interface as many other Microsoft products. Rather
than entering a URL to a SharePoint or OneDrive location, you can select one of your
sites by using the Quick access section or selecting More places.

If you don't have a subscription, the OneDrive for Business and SharePoint options are
unavailable, but you can still select Browse to get local files from your computer. This
image shows the unavailable options, but the Browse option is enabled:
You can't use Upload to get files from personal OneDrive accounts, but you can upload
files from your computer.

Import Excel data into Power BI

To import Excel data into Power BI, in My workspace, select New > Semantic model >
Excel, and then find the file.

The My files list allows you to add files from your documents folder and other personal
sources.

You can use the Quick access list on the left side of the window to add files from
SharePoint sites and other shared sources.

Select Browse this device to add files from the device you're currently using.

When you import Excel data, Power BI imports any supported data in tables and any
data model into a new Power BI semantic model.

You should import your data if you used Get & Transform Data or Power Pivot to load
data into a data model.

If you upload from OneDrive for Business, when you save changes, Power BI
synchronizes them with the semantic model in Power BI, usually within about an hour.
You can also select Publish to export your changes immediately. Any visualizations in
reports and dashboards also update, based on the following refresh triggers:

Report tiles Dashboard tiles

Open the report, after the Open the dashboard, after the cache refreshes.
cache expires.

Select Refresh in the report. Select Refresh in the dashboard.

 Report tiles Dashboard tiles

Automatically for pinned tiles when the cache refreshes, if the

dashboard is already open.

７ Note

Pinned report pages don't support the automatic refresh feature.

Prepare your workbook for Power BI

Watch this video to learn more about how to make sure your Excel workbooks are ready
for Power BI:

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI service.

https://www.youtube-nocookie.com/embed/l2wy4XgQIu0

Where to save your workbook file

Where you save your workbook file makes a difference.

Local. If you save your workbook file to a drive on your computer or another
location in your organization, you can load your file into Power BI. Your file actually
remains on the source drive. When you import the file, Power BI creates a new
semantic model and loads data and any data model from the workbook into the
semantic model.

Local Excel workbooks open in Excel Online within Power BI. Unlike Excel
workbooks stored on OneDrive or SharePoint team sites, you can't edit local Excel
files within Power BI.

Excel also has a Publish command on the File menu. Using this Publish command
is effectively the same as using Upload > Browse from Power BI. If you regularly
make changes to the workbook, it's often easier to update your semantic model in
Power BI.

OneDrive for Business. Signing in to OneDrive for Business with the same account
as Power BI is the most effective way to keep your work in Excel in sync with your
 Power BI semantic model, reports, and dashboards. Both Power BI and OneDrive
are in the cloud, and Power BI connects to your workbook file on OneDrive about
once an hour. If Power BI finds any changes, it automatically updates your Power BI
semantic model, reports, and dashboards.

As when you have a file saved to a local drive, you can use Publish in Excel to
update your Power BI semantic model and reports immediately. Otherwise, Power
BI automatically synchronizes, usually within an hour.

SharePoint team site. Saving your Power BI Desktop files to a SharePoint team site
is almost the same as saving them to OneDrive for Business. The biggest difference
is how you connect to the file from Power BI. You can specify a URL or connect to
the root folder.

Publish from Excel to your Power BI site

Using the Excel Publish to Power BI feature is effectively the same as using Power BI to
import or connect to your file. For more information, see Publish to Power BI from
Microsoft Excel.

７ Note

If you upload an Excel workbook that's connected to an on-premises SQL Server

Analysis Services (SSAS) cube, you can't refresh the underlying data model in the
Power BI service.

Migrating your Excel workbooks

For local Excel workbooks uploaded to a Power BI workspace, use the Download Excel
file option to download the workbook. Then save it to OneDrive for Business or a
SharePoint Document library (ODSP). You can then import the workbook from ODSP to
the workspace again.
To refresh data in Excel data models, you'll need to publish the data model as a Power BI
semantic model. We recommend using the Power BI Desktop to import the model
because it upgrades your data model to the latest version. This gives you the best future
experience. Use the Import from Power Query, Power Pivot, Power View option on
Power BI Desktop's File menu.

To build new workbooks connected to a semantic data model in your Excel workbook,
you should first publish the data model as a Power BI semantic model. Then in Excel use
the From Power BI (Microsoft) option to connect your workbook to the semantic
model. This option is available in the Data ribbon, under Get Data in the From Power
Platform menu.

For cases where you include a workbook in a Power BI organizational app, remember to
republish the app with the new items.

To learn which workbooks can be affected by the deprecation of local workbooks and
refresh capabilities, use the workbooks Power BI admin REST API. It lists the workbooks
in your organization. You must be a member of the Power BI admin role or a Global
Administrator to call this API.

GET https://api.powerbi.com/v1.0/myorg/admin/workbooks
The API provides a list of all the Excel workbooks published in your organization. The list
is formatted in JSON.

Below is an example output for the API.

[
{
"DisplayName": "Workbook without a Data Model",
"WorkspaceName": "My workspace",
"HasDataModel": false,
"HasScheduledRefreshOnDataModel": false,
"UploadedOn": "2023-07-28T10:54:17.093"
},
{
"DisplayName": "Workbook with Data Model",
"WorkspaceName": "My workspace",
"HasDataModel": true,
"HasScheduledRefreshOnDataModel": true,
"UploadedBy": "[email protected]",
"UploadedOn": "2022-11-16T09:51:17.497"
}
]

You can check if the Excel workbook is a local workbook by navigating to it in Power BI
and seeing if it has the Download Excel file option is available.

You can use PowerShell to call the API as shown in the example below:

Invoke-PowerBIRestMethod -Url
"https://api.powerbi.com/v1.0/myorg/admin/workbooks" -Method GET

To use PowerShell, first install the required MicrosoftPowerBIMgmt module. See Power
BI Cmdlets reference for details. You will need to call Login-PowerBIServiceAccount
commandlet before calling Invoke-PowerBIRestMethod.

Troubleshooting and limitations

If your workbook file is too large, see Reduce the size of an Excel workbook to view
it in Power BI.

The upload of Excel workbooks to a Power BI workspace isn't supported for

sovereign cloud customers.
 You can't use scheduled refresh for Excel workbooks that have connections to on-
premises SSAS tabular models through a gateway.

Next steps
Explore your data. After you upload data and reports from your file into Power BI,
you can select the new semantic model to explore the data. When you select the
workbook, it opens in Power BI the same as if it were in Excel Online.

Schedule refresh. If your Excel workbook connects to external data sources, or if

you imported from a local drive, you can set up scheduled refresh to make sure
your semantic model or report is always up-to-date. In most cases, setting up
scheduled refresh is easy to do. For more information, see Data refresh in Power BI.

Publish to Power BI from Microsoft Excel.

Get data from Power BI Desktop files
Article • 11/10/2023

Power BI Desktop makes business intelligence and reporting easy. Whether you're
connecting to many different data sources, querying and transforming data, modeling
your data, and creating powerful and dynamic reports, Power BI Desktop makes
business intelligence tasks intuitive and fast. If you're not familiar with Power BI Desktop,
check out Getting started with Power BI Desktop.

Once you bring data into Power BI Desktop and create a few reports, it's time to get
your saved file into the Power BI service.

Where your file is saved makes a difference

There are several locations where you might store Power BI Desktop files:

Local. If you save your file to a local drive on your computer or another location in
your organization, you can import your file, or you can publish from Power BI
Desktop to get its data and reports into the Power BI service.

Your file remains on your local drive. The whole file isn't moved into Power BI. A
new semantic model is created in Power BI and data and the data model from the
Power BI Desktop file are loaded into the semantic model. If your file has any
reports, those reports appear in your Power BI service site under Reports.

OneDrive for work or school. By far, the most effective way to keep your work in
Power BI Desktop in sync with the Power BI service is to use your OneDrive for
work or school and sign in with the same account as the Power BI service. Your
work includes semantic model, reports, and dashboards. Because both the Power
BI service and OneDrive are in the cloud, Power BI connects to your file on
OneDrive about every hour. If it finds any changes, your semantic model, reports,
and dashboards are updated in the Power BI service.

OneDrive - Personal. If you save your files to your own OneDrive account, you get
many of the same benefits as you would with OneDrive for work or school. The
biggest difference is when you first connect to your file, you need to sign in to
 your OneDrive with your Microsoft account. This account is usually different from
what you use to sign in to the Power BI service.

When signing in with your OneDrive with your Microsoft account, be sure to select
the Keep me signed in option. This way, the Power BI service can connect to your
file about every hour and make sure that your semantic model in the Power BI
service is in-sync.

SharePoint Team-Sites. Saving your Power BI Desktop files to SharePoint – Team

Sites is much the same as saving to OneDrive for work or school. The biggest
difference is how you connect to the file from the Power BI service. You can specify
a URL or connect to the root folder. You can also set up a Sync folder that points
to the SharePoint folder. Files in that folder sync up with the ones on SharePoint.

Streamlined upload to Power BI

Beginning in November 2022, there's a new and streamlined experience for uploading
files to the Power BI service. In the workspace into which you want to add files, you see
an Upload dropdown menu option next to the New button. You can use the dropdown
menu to connect to files stored in OneDrive for work or school or any SharePoint site to
which you have access, or you can upload them from your computer through the Browse
menu option. The following image shows the menu options.

If you choose to upload a local file, a copy of the file is added to the workspace. If you
use the OneDrive for work or school or SharePoint option, the Power BI service creates a
connection to the file and as you make changes to the file in SharePoint, Power BI can
automatically sync those changes approximately each hour.
A benefit of uploading files this way, in addition to being easy to use, is that the
OneDrive for work or school and SharePoint options use the same file selection interface
used in many other Microsoft products.

Rather than having to paste a direct URL to a given SharePoint site, which was
previously required, you can now simply select one of your sites through the Quick
access section or the More places links.

When you upload an Excel file this way, your workbook appears in the Power BI service
just like it would in Excel Online, as shown in the following image.

If you don't have a subscription, OneDrive for work or school and SharePoint options are
disabled, but you can still browse for local files on your computer. The following image
shows the subscription options disabled, with the Browse option highlighted.

７ Note

You can't upload files from SharePoint Document set folder or from personal
OneDrive accounts.

Publish a file from Power BI Desktop to the

Power BI service
Using Publish from Power BI Desktop is similar uploading files in the Power BI service.
Both initially import your file data from a local drive or connect to it on OneDrive.
However, there are differences. If you upload from a local drive, refresh that data
frequently to ensure the online and local copies of the data are current with each other.
Here's the quick how to, but you can see Publish from Power BI Desktop to learn more.

1. In Power BI Desktop, select File > Publish > Publish to Power BI, or select Publish
on the ribbon.

2. Sign in to the Power BI service. You only need to sign in the first time.

When complete, you get a link to open your report in your Power BI site.

Next steps
Explore your data: Once you get data and reports from your file into the Power BI
service, it's time to explore. If your file already has reports in it, they appear in the
navigator pane in Reports. If your file just had data, you can create new reports;
just right-click the new semantic model and then select Explore.

Refresh external data sources: If your Power BI Desktop file connects to external
data sources, you can set up scheduled refresh to make sure your semantic model
is always up-to-date. In most cases, setting up scheduled refresh is easy to do, but
going into the details is outside the scope of this article. See Data refresh in Power
BI to learn more.
Edit parameter settings in the Power BI
service
Article • 11/10/2023

Report creators add query parameters to reports in Power BI Desktop. Parameters allow
you to make parts of reports depend on one or more parameter values. For example, a
report creator might create a parameter that restricts the data to a single
country/region, or a parameter that defines acceptable formats for fields like dates,
time, and text.

Review and edit parameters in Power BI service

As a report creator, you define parameters in Power BI Desktop. When you publish that
report to the Power BI service, the parameter settings and selections travel with it. You
can review and edit parameter settings in the Power BI service, but not create them.

1. In the Power BI service, select the cog icon and then choose Settings.

2. Select the tab for Semantic models and highlight a semantic model in the list.
3. Expand Parameters. If the selected semantic model has no parameters, you see a
message with a link to Learn more about query parameters. If the semantic model
does have parameters, expand the Parameters heading to reveal those
parameters.

4. Review the parameter settings and make changes if needed.

Considerations and limitations
Grayed-out fields aren't editable. Any and Binary type parameters work in Power BI
Desktop. The Power BI service doesn't currently support them for security reasons.

Next steps
An ad-hoc way to add simple parameters is by modifying filters.
Get data from comma separated value
(CSV) files
Article • 01/17/2024

Comma separated value files, often known as a CSV, are simple text files with rows of
data where each value is separated by a comma. These types of files can contain large
amounts of data within a relatively small file size, making them an ideal data source for
Power BI. You can download a sample CSV file .

If you have a CSV, it’s time to get it into your Power BI site as a semantic model where
you can begin exploring your data, create some dashboards, and share your insights
with others.

 Tip

Many organizations output a CSV with updated data each day. To make sure your
semantic model in Power BI stays in-sync with your updated file, be sure the file is
saved to OneDrive with the same name.

Where your file is saved makes a difference

Local - If you save your CSV file to a local drive on your computer or another location in
your organization, you can import your file into Power BI. Your file will actually remain on
your local drive, so the whole file isn’t imported into Power BI. What really happens is a
new semantic model is created in Power BI and data from the CSV file is loaded into the
semantic model.

OneDrive for work or school – If you have OneDrive for work or school and you sign
into it with the same account you use to sign into Power BI, this method is the most
effective way to keep your CSV file and your semantic model, reports, and dashboards in
Power BI in-sync. Because both Power BI and OneDrive are in the cloud, Power BI
connects to your file on OneDrive about every hour. If any changes are found, your
semantic model, reports, and dashboards are automatically updated in Power BI.
OneDrive - Personal – If you save your files to your own OneDrive account, you’ll get
many of the same benefits as you would with OneDrive for work or school. The biggest
difference is when you first connect to your file you’ll need to sign in to your OneDrive
with your Microsoft account, which is different from what you use to sign in to Power BI.
When signing into your OneDrive with your Microsoft account, be sure to select the
Keep me signed in option. This way, Power BI will be able to connect to your file about
every hour and make sure your semantic model in Power BI is in-sync.

SharePoint – Saving your Power BI Desktop files to SharePoint is much the same as
saving to OneDrive for work or school. The biggest difference is how you connect to the
file from Power BI. You can specify a URL or connect to the root folder.

Import or connect to a CSV file

） Important

The maximum file size you can import into Power BI is 1 GB.

1. In a Power BI workspace, select My workspace > New > Semantic model.

2. In the window that appears, select CSV.

 3. Go to the file you want to upload and then choose Import. A new Semantic model
details window appears in the main pane of Power BI.

Next steps
Explore your data - Once you get data from your file into Power BI, it's time to explore.
Select More options (...), and then choose an option from the menu.
Schedule refresh - If your file is saved to a local drive, you can schedule refreshes so
your semantic model and reports in Power BI stay up-to-date. To learn more, see Data
refresh in Power BI. If your file is saved to OneDrive, Power BI will automatically
synchronize with it about every hour.
Real-time streaming in Power BI
Article • 11/10/2023

Power BI with real-time streaming helps you stream data and update dashboards in real
time. Any visual or dashboard created in Power BI can display and update real-time data
and visuals. The devices and sources of streaming data can be factory sensors, social
media sources, service usage metrics, or many other time-sensitive data collectors or
transmitters.

This article shows you how to set up and use real-time streaming semantic models in
Power BI.

Types of real-time semantic models

First, it's important to understand the types of real-time semantic models that are
designed to display in tiles and dashboards, and how those semantic models differ.

The following three types of real-time semantic models are designed for display on real-
time dashboards:

Push semantic model

 Streaming semantic model
PubNub streaming semantic model

This section explains how these semantic models differ from one another. Later sections
describe how to push data into each of these semantic models.

Push semantic model

With a push semantic model, data is pushed into the Power BI service. When the
semantic model is created, the Power BI service automatically creates a new database in
the service to store the data.

Because there's an underlying database that stores the data as it arrives, you can create
reports with the data. These reports and their visuals are just like any other report
visuals. You can use all of Power BI's report building features, such as Power BI visuals,
data alerts, and pinned dashboard tiles.

Once you create a report using the push semantic model, you can pin any of the report
visuals to a dashboard. On that dashboard, visuals update in real time whenever the
data is updated. Within the Power BI service, the dashboard triggers a tile refresh every
time new data is received.

There are two considerations to note about pinned tiles from a push semantic model:

Pinning an entire report by using the Pin live option won't result in the data
automatically being updated.
Once you pin a visual to a dashboard, you can use Q&A to ask questions about the
push semantic model in natural language. After you make a Q&A query, you can
pin the resulting visual back to the dashboard, and that visual will also update in
real time.

Streaming semantic model

A streaming semantic model also pushes data into the Power BI service, with an
important difference: Power BI stores the data only into a temporary cache, which
quickly expires. The temporary cache is used only to display visuals that have some
transient history, such as a line chart that has a time window of one hour.

A streaming semantic model has no underlying database, so you can't build report
visuals by using the data that flows in from the stream. Therefore, you can't use report
functionality such as filtering, Power BI visuals, and other report functions.
The only way to visualize a streaming semantic model is to add a tile and use the
streaming semantic model as a custom streaming data source. The custom streaming
tiles that are based on a streaming semantic model are optimized for quickly displaying
real-time data. There's little latency between pushing the data into the Power BI service
and updating the visual, because there's no need for the data to be entered into or read
from a database.

In practice, it's best to use streaming semantic models and their accompanying
streaming visuals in situations when it's critical to minimize the latency between pushing
and visualizing data. You should have the data pushed in a format that can be visualized
as-is, without any more aggregations. Examples of data that's ready as-is include
temperatures and pre-calculated averages.

PubNub streaming semantic model

With a PubNub streaming semantic model, the Power BI web client uses the PubNub
SDK to read an existing PubNub data stream. The Power BI service stores no data.
Because the web client makes this call directly, if you allow only approved outbound
traffic from your network, you must list traffic to PubNub as allowed. For instructions,
see the support article about approving outbound traffic for PubNub .

As with the streaming semantic model, with the PubNub streaming semantic model
there's no underlying Power BI database. You can't build report visuals against the data
that flows in, and can't use report functionality like filtering or Power BI visuals. You can
visualize a PubNub streaming semantic model only by adding a tile to the dashboard
and configuring a PubNub data stream as the source.

Tiles based on a PubNub streaming semantic model are optimized for quickly displaying
real-time data. Since Power BI is directly connected to the PubNub data stream, there's
little latency between pushing the data into the Power BI service and updating the
visual.

Streaming semantic model matrix

The following table describes the three types of semantic models for real-time
streaming and lists their capabilities and limitations.

Capability Push Streaming PubNub

Dashboard tiles Yes. Yes. Yes.

update in real time as For visuals built via For custom streaming For custom streaming
data is pushed in
 Capability Push Streaming PubNub

reports and then tiles added directly to tiles added directly to

pinned to dashboard. the dashboard. the dashboard.

Dashboard tiles No. Yes. Yes.

update with smooth
animations

Data stored Yes. No. No.

permanently in Power Data is temporarily
BI for historic analysis stored for one hour to
render visuals.

Build Power BI reports Yes. No. No.

on top of the data

Max rate of data 1 requests 5 requests N/A

ingestion 16 MB/request 15 KB/request Data isn't being
pushed into Power BI.

Limits on data 1M rows/hour None. N/A

throughput Data isn't being
pushed into Power BI.

Push data to semantic models

This section describes how to create and push data into the three primary types of real-
time semantic models that you can use in real-time streaming.

You can push data into a semantic model by using the following methods:

The Power BI REST APIs

The Power BI streaming semantic model UI
Azure Stream Analytics

Use Power BI REST APIs to push data

You can use Power BI REST APIs to create and send data to push semantic models and to
streaming semantic models. When you create a semantic model by using Power BI REST
APIs, the defaultMode flag specifies whether the semantic model is push or streaming.

If no defaultMode flag is set, the semantic model defaults to a push semantic model. If
the defaultMode value is set to pushStreaming , the semantic model is both a push and
streaming semantic model, and provides the benefits of both semantic model types.
 ７ Note

When you use semantic models with the defaultMode flag set to pushStreaming , if a
request exceeds the 15 KB size restriction for a streaming semantic model, but is
less than the 16 MB size restriction for a push semantic model, the request
succeeds and the data updates in the push semantic model. However, any
streaming tiles temporarily fail.

Once a semantic model is created, you can use the PostRows REST APIs to push data. All
requests to REST APIs are secured by using Azure Active Directory (Azure AD) OAuth.

Use the streaming semantic model UI to push data

In the Power BI service, you can create a semantic model by selecting the API approach,
as shown in the following screenshot:
When you create the new streaming semantic model, you can enable Historic data
analysis, as shown in the following screenshot. This selection has a significant impact.
When Historic data analysis is disabled, as it is by default, you create a streaming
semantic model as described earlier. When Historic data analysis is enabled, the
semantic model you create becomes both a streaming semantic model and a push
semantic model. This setting is equivalent to using the Power BI REST APIs to create a
semantic model with its defaultMode set to pushStreaming , as described earlier.

７ Note

Streaming semantic models created by using the Power BI service UI don't require
Azure AD authentication. In such semantic models, the semantic model owner
receives a URL with a rowkey, which authorizes the requestor to push data into the
semantic model without using an Azure AD OAuth bearer token. However, the
Azure AD approach still works to push data into the semantic model.

Use Azure Stream Analytics to push data

You can add Power BI as an output within Azure Stream Analytics, and then visualize
those data streams in the Power BI service in real time. This section describes the
technical details of that process.

Azure Stream Analytics uses the Power BI REST APIs to create its output data stream to
Power BI, with defaultMode set to pushStreaming . The resulting semantic model can use
both push and streaming. When you create the semantic model, Azure Stream Analytics
sets the retentionPolicy flag to basicFIFO . With that setting, the database that
supports the push semantic model stores 200,000 rows, and drops rows in a first-in first-
out (FIFO) fashion.

） Important

If your Azure Stream Analytics query results in very rapid output to Power BI, for
example once or twice per second, Azure Stream Analytics begins batching the
outputs into a single request. This batching might cause the request size to exceed
the streaming tile limit, and streaming tiles might fail to render. In this case, the
best practice is to slow the rate of data output to Power BI. For example, instead of
a maximum value every second, request a maximum value over 10 seconds.

Set up your real-time streaming semantic

model in Power BI
To get started with real-time streaming, you choose one of the following ways to
consume streaming data in Power BI:

Tiles with visuals from streaming data

Semantic models created from streaming data that persist in Power BI

For either option, you need to set up streaming data in Power BI. To get your real-time
streaming semantic model working in Power BI:

1. In either an existing or new dashboard, select Add a tile.

2. On the Add a tile page, select Custom Streaming Data, and then select Next.
3. On the Add a custom streaming data tile page, you can select an existing
semantic model, or select Manage semantic models to import your streaming
semantic model if you already created one. If you don't have streaming data set up
yet, select Add streaming semantic model to get started.
4. On the New streaming semantic model page, select API, Azure Stream, or
PubNub, and then select Next.
Create a streaming semantic model
There are three ways to create a real-time streaming data feed that Power BI can
consume and visualize:

Power BI REST API using a real-time streaming endpoint

Azure Stream
PubNub

This section describes the Power BI REST API and PubNub options, and explains how to
create a streaming tile or semantic model from the streaming data source. You can then
use the semantic model to build reports. For more information about the Azure Stream
option, see Power BI output from Azure Stream Analytics.
Use the Power BI REST API
The Power BI REST API makes real-time streaming easier for developers. After you select
API on the New streaming semantic model screen and select Next, you can provide
entries that enable Power BI to connect to and use your endpoint. For more information
about the API, see Use the Power BI REST APIs.

If you want Power BI to store the data this data stream sends, so you can do reporting
and analysis on the collected data, enable Historic data analysis.

After you successfully create your data stream, you get a REST API URL endpoint. Your
application can call the endpoint by using POST requests to push your streaming data to
the Power BI semantic model. In your POST requests, ensure that the request body
matches the sample JSON that the Power BI user interface provided. For example, wrap
your JSON objects in an array.

Ｕ Caution

For streaming semantic models you create in the Power BI service UI, the semantic
model owner gets a URL that includes a resource key. This key authorizes the
requestor to push data into the semantic model without using an Azure AD OAuth
bearer token. Keep in mind the implications of having a secret key in the URL when
you work with this type of semantic model and method.

Use PubNub
The integration of PubNub streaming with Power BI helps you create and use your low-
latency PubNub data streams in Power BI. When you select PubNub on the New
streaming semantic model screen and select Next, you see the following screen:
 ） Important

You can secure PubNub channels by using a PubNub Access Manager (PAM)
authentication key. This key is shared with all users who have access to the
dashboard. For more information about PubNub access control, see Manage
Access .

PubNub data streams are often high volume, and aren't always suitable for storage and
historical analysis in their original form. To use Power BI for historical analysis of PubNub
data, you must aggregate the raw PubNub stream and send it to Power BI, for example
by using Azure Stream Analytics .
Example of real-time streaming in Power BI
Here's an example of how real-time streaming in Power BI works. This sample uses a
publicly available stream from PubNub. Follow along with the example to see the value
of real-time streaming for yourself.

1. In the Power BI service, select or create a new dashboard. At the top of the screen,
select Edit > Add a tile.

2. On the Add a tile screen, select Custom Streaming Data, and then select Next.
3. On the Add a custom streaming data tile page, select Add streaming semantic
model.
4. On the New streaming semantic model page, select PubNub, and then select
Next.

5. On the next screen, enter a Semantic model name, enter the following values into
the next two fields, and then select Next.

Sub-key: sub-c-99084bc5-1844-4e1c-82ca-a01b18166ca8
Channel name: pubnub-sensor-network
6. On the next screen, keep the automatically populated values, and select Create.
7. Back in your Power BI workspace, create a new dashboard, and at the top of the
screen, select Edit > Add a tile.

8. Select Custom Streaming Data, and select Next.

9. On the Add a custom streaming data tile page, select your new streaming
semantic model, and then select Next.
 Play around with the sample semantic model. By adding value fields to line charts
and adding other tiles, you can get a real-time dashboard that looks like the
following screenshot:

Go on to create your own semantic models and stream live data to Power BI.

Questions and answers

Here are some common questions and answers about real-time streaming in Power BI.

Can you use filters on push or streaming semantic models?

Streaming semantic models don't support filtering. For push semantic models, you can
create a report, filter the report, and then pin the filtered visuals to a dashboard.
However, there's no way to change the filter on the visual once it's on the dashboard.

You can pin the live report tile to the dashboard separately, and then you can change
the filters. However, live report tiles won't update in real time as data is pushed in. You
have to manually update the visual by selecting the Refresh icon at top right on the
dashboard page.

When you apply filters to push semantic models that have DateTime fields with
millisecond precision, equivalence operators aren't supported. Operators such as greater
than > or less than < operate properly.

How do you see the latest value on push or streaming semantic

models?

Streaming semantic models are designed to display the latest data. You can use the
Card streaming visual type to easily see the latest numeric values. Card visuals don't
support DateTime or Text data types.

For push semantic models, if you have a timestamp in the schema, you can try creating
a report visual with the last N filter.

How can you do modeling on real-time semantic models?

Modeling isn't possible on a streaming semantic model, because the data isn't stored
permanently. For a push semantic model, you can use the create semantic model REST
API to create a semantic model with relationship and measures, and use the update
table REST APIs to add measures to existing tables.

How can you clear all the values on a push or streaming semantic
model?
On a push semantic model, you can use the delete rows REST API call. There's no way to
clear data from a streaming semantic model, although the data will clear itself after an
hour.

If you set up an Azure Stream Analytics output to Power BI but you

don't see it in Power BI, what's wrong?
Take these steps to troubleshoot the issue:

1. Restart the Azure Stream Analytics job.

2. Try reauthorizing your Power BI connection in Azure Stream Analytics.
3. Make sure that you're checking the same workspace in the Power BI service that
you specified for the Azure Stream Analytics output.
4. Make sure the Azure Stream Analytics query explicitly outputs to the Power BI
output by using the INTO keyword.
5. Determine whether the Azure Stream Analytics job has data flowing through it. The
semantic model is created only when data is being transmitted.
6. Look into the Azure Stream Analytics logs to see if there are any warnings or
errors.
Automatic page refresh
You can use automatic page refresh at a report page level to set a refresh interval for
visuals that's only active when the page is being consumed. Automatic page refresh is
available only for DirectQuery data sources. The minimum refresh interval depends on
the type of workspace where the report is published and capacity admin settings for
Premium workspaces.

For more information about automatic page refresh, see Automatic page refresh in
Power BI.

Next steps
Overview of the Power BI REST API with real-time data
Load data in a Power BI streaming semantic model and build a dataflows
monitoring report with Power BI
Publish to Power BI from Microsoft
Excel
Article • 11/10/2023

） Important

The following capabilities are deprecated and will no longer be available starting
September 29th, 2023:

Upload of local workbooks to Power BI workspaces will no longer be allowed.

Configuring scheduling of refresh and refresh now for Excel files that don’t
already have scheduled refresh configured will no longer be allowed.

The following capabilities are deprecated and will no longer be available starting
October 31, 2023:

Scheduled refresh and refresh now for existing Excel files that were previously
configured for scheduled refresh will no longer be allowed.
Local workbooks uploaded to Power BI workspaces will no longer open in
Power BI.

After October 31, 2023:

You can download existing local workbooks from your Power BI workspace.
You can publish your Excel data model as a Power BI semantic model and
schedule refresh.
You can import Excel workbooks from OneDrive and SharePoint Document
libraries to view them in Power BI.

If your organization uses these capabilities, see more details in Migrating your
Excel workbooks.

With Microsoft Excel 2016 and later, you can publish your Excel workbooks directly to
your Power BI workspace. In Power BI, you can create highly interactive reports and
dashboards based on your workbook data. You can then share your insights with others
in your organization.
When you publish a workbook to Power BI, there are few things to consider:

You must use the same account to sign in to Office, OneDrive for work or school if
your workbooks are saved there, and Power BI.
You can't publish an empty workbook, or a workbook that doesn't have any Power
BI supported content.
You can't publish encrypted or password protected workbooks, or workbooks with
Information Protection Management applied.
Publishing to Power BI requires modern authentication to be enabled, the default.
Otherwise, the Publish option isn't available from the File menu.
Publishing to Power BI from Excel Desktop isn't supported for sovereign clouds.

Publish your Excel workbook

To publish your Excel workbook to Power BI, in Excel, select File > Publish and select
either Upload or Export. The following screenshot shows the two options for how to get
your workbook into Power BI:
 If you select Upload, you can interact with the workbook just as you would in Excel
Online. You can also pin selections from your workbook onto Power BI dashboards,
and share your workbook or selected elements through Power BI.

If you select Export, you can export table data and its data model into a Power BI
semantic model, and use the semantic model to create Power BI reports and
dashboards.

When you select Publish, you can select the workspace to publish to. If your Excel file is
on OneDrive for work or school, you can publish only to your My Workspace. If your
Excel file is on a local drive, you can publish to My Workspace or to a shared workspace
you can access.

Publish local files

Excel supports publishing local Excel files. Files don't need to be saved to OneDrive for
work or school or to SharePoint Online.

） Important
 You can publish local files only if you're using Excel 2016 or later with a Microsoft
365 subscription. Excel 2016 standalone installations can publish to Power BI, but
only when the workbook is saved to OneDrive for work or school or to SharePoint
Online.

Once published, the workbook content you publish imports into Power BI, separate from
the local file. If you want to update the file in Power BI, you must update the local file
and publish the updated version again. Or, you can refresh the data by configuring
scheduled refresh on the workbook or semantic model in Power BI.

Publish from a standalone Excel installation

When you publish from a standalone Excel installation, you must save the workbook to
OneDrive for work or school. Select Save to Cloud and choose a location in OneDrive
for work or school.

Once you save your workbook to OneDrive for work or school, when you select Publish,
you can use the Upload or Export options to get your workbook into Power BI.
Upload your workbook to Power BI

When you choose the Upload option, your workbook appears in Power BI just as it
would in Excel Online. But unlike in Excel Online, you have options to let you pin
elements from your worksheets to dashboards.

If you choose Upload, you can't edit your workbook in Power BI. If you need to change
the data, you can select Edit, and then choose to edit your workbook in Excel Online or
open it in Excel on your computer. Any changes you make are saved to the workbook
on OneDrive for work or school.

When you choose Upload, no semantic model is created in Power BI. Your workbook
appears in your workspace navigation pane under Reports. Workbooks uploaded to
Power BI have an Excel icon that identifies them as uploaded Excel workbooks.

Choose the Upload option if you only have data in worksheets, or you have PivotTables
and Charts you want to see in Power BI.

Using Upload from Publish to Power BI in Excel is a similar experience to using Upload
> OneDrive for Business > Upload in Power BI, and then opening the file in Excel Online
from Power BI in your browser.

Export workbook data to Power BI

When you choose the Export option, any supported data in tables and/or a data model
are exported into a new semantic model in Power BI. You can continue editing your
workbook. When you save your changes, they synchronize with the semantic model in
Power BI, usually within about an hour. If you need more immediate updates, you can
select Publish again from Excel to export your changes immediately. Any visualizations
in reports and dashboards update too.

Choose the Export option if you used the Get & Transform data or Power Pivot features
to load data into a data model.

Using Export is similar to using New > Upload a file > Excel > Import from Power BI in
your browser.

Publish
When you choose either Upload or Export, Excel signs in to Power BI with your current
account and publishes your workbook to your Power BI workspace. You can monitor the
status bar in Excel to see publishing progress.

When publishing is complete, you can go to Power BI directly from Excel.

Next steps
Excel data in Power BI
More questions? Try the Power BI Community.
Reduce the size of an Excel workbook to
view it in Power BI
Article • 02/09/2023

You can upload any Excel workbook smaller than 1 GB to Power BI. An Excel workbook
can have two parts: a Data Model, and the rest of the report—the core worksheet
contents. If the report meets the following size limits, you can save it to OneDrive for
work or school, connect to it from Power BI, and view it in Excel Online:

The workbook can be up to 1 GB.

The core worksheet contents can be up to 30 MB.

What makes core worksheet contents larger

than 30 MB
Here are some elements that can make the core worksheet contents larger than 30 MB:

Images
Shaded cells
Colored worksheets
Text boxes
Clip art

Consider removing these elements, if possible.

If the report has a Data Model, you have some other options:

Remove data from Excel worksheets and store it in the Data Model. For more
information, see the Remove data from worksheets section.
Create a memory-efficient Data Model to reduce the overall size of the report.

To make any of these changes, you need to edit the workbook in Excel.

For more information, see File size limits for Excel workbooks in SharePoint Online .

Remove data from worksheets

If you import data into Excel from the Power Query Editor or the Excel Data tab, the
workbook might have the same data in an Excel table and in the Data Model. Large
tables in Excel worksheets might make the core worksheet contents more than 30 MB.
Removing the table in Excel and keeping the data in the Data Model can greatly reduce
the core worksheet contents of the report.

When you import data into Excel, follow these tips:

In Power Query Editor: Clear the Load to worksheet box under File > Options and
settings > Query Options.

This imports data only into the Data Model, with no data in Excel worksheets.

From the Excel Data tab, if you previously selected Table in the import wizard:

1. Go to Existing Connections.

2. Select the connection, and then select Open. Select Only Create Connection.
 3. Delete the original table or tables created during the initial import.

Workbook size optimizer

If your workbook contains a data model, you can run the workbook size optimizer to
reduce the size of your workbook. For more information, see Download Workbook Size
Optimizer .

Related info
Create a memory-efficient Data Model by using Excel and the Power Pivot add-in

Use OneDrive for work or school links in Power BI Desktop

Introduction to semantic models across
workspaces
Article • 11/10/2023

Business intelligence is a collaborative activity. It's important to establish standardized

semantic models that can be the one source of truth. Then discovering and reusing those
standardized semantic models is key. When expert data modelers in your organization
create and share optimized semantic models, report creators can start with those
semantic models to build accurate reports. Then your organization has consistent data
for making decisions, and a healthy data culture.

In Power BI, semantic model creators can control who has access to their data by using
the Build permission. Semantic model creators can also certify or promote semantic
models so others can discover them. That way, report authors know which semantic
models are high quality and official, and they can use those semantic models wherever
they author in Power BI. Administrators have a new tenant setting to govern the use of
semantic models across workspaces.

Semantic model sharing and workspaces

Building reports based on semantic models in different workspaces, and copying reports
to different workspaces, are tightly coupled with the workspace:

In the Power BI service, when you open the semantic model catalog from a
workspace, the semantic model catalog shows semantic models in your My
workspace and in other workspaces.
In Power BI Desktop, you can publish Live Connect reports to different workspaces.

Discover semantic models

When you build a report on top of an existing semantic model, the first step is to
connect to the semantic model, either in the Power BI service or Power BI Desktop. Read
about discovering semantic models from different workspaces

Copy a report
When you find a report you like, in a workspace or an app, you can make a copy of it,
and then modify it to fit your needs. You don't have to worry about creating the data
model. The data model is already created for you. And it's much easier to modify an
existing report than it is to start from scratch. Read more about copying reports.

Build permission for semantic models

With Build permission type, if you're a semantic model creator, you can determine who
in your organization can build new content on your semantic models. People with Build
permission can also build new content on the semantic model outside Power BI, such as
Excel sheets via Analyze in Excel, XMLA, and export. Read more about the Build
permission.

Promotion and certification

If you create semantic models, when you create one that others can benefit from, you
can make it easier for them to discover it by promoting your semantic model. You can
also request that experts in your organization certify your semantic model.

Licensing
The specific features and experiences built on shared semantic model capabilities are
licensed according to their existing scenarios. For example:

In general, discovering and connecting to shared semantic models is available to

anyone. It isn't a feature restricted to Premium.
Users without a Pro or Premium Per User (PPU) license can only use semantic
models across workspaces for report authoring if those semantic models reside in
the users' personal My workspace or in a Premium-backed workspace. The same
licensing restriction applies whether they author reports in Power BI Desktop or in
the Power BI service.
Copying reports between workspaces requires a Pro or Premium Per User license.
Copying reports from an app requires a Pro or Premium Per User license.
 Promoting and certifying semantic models requires a Pro or Premium Per User
license.

Considerations and limitations

As an app publisher, you have to make sure that your audience has access to
semantic models outside of the workspace. Otherwise, users will encounter issues
when interacting with your app: reports won't open without semantic model access
and dashboard tiles will show as locked. Also, users won't able to open the app if
the first item in its navigation is a report without access to the semantic model.
By design, Publish to web doesn't work for a report based on a shared semantic
model.
If two people are members of a workspace that is accessing a shared semantic
model, it's possible that only one of them can see the related semantic model in
the workspace. Only people with at least Read access to the semantic model can
see the shared semantic model.

Next steps
Promote semantic models
Certify semantic models
Request semantic model certification
Govern the use of semantic models across workspaces
Questions? Try asking the Power BI Community
Semantic model description
Article • 11/10/2023

To help members of your organization quickly identify semantic models that might be
useful for them, provide a concise, informative description of your semantic model in
the semantic model's settings. Users see this description in the tooltip next to the
semantic model's name in the semantic models hub and on the semantic model's
details page.

Providing a meaningful description helps foster semantic model reuse. For instance,
based on a semantic model's description, users may decide to explore reports that are
based on the semantic model, or to create their own reports based on the semantic
model.

Provide a semantic model description

To provide a description for a semantic model, go to semantic model's settings page,
find the Semantic model description section, and enter your description in the text box.

In the settings page, find the Semantic model description section, and enter your
description in the text box. Select Apply to save your description.

Next steps
OneLake data hub
Questions? Try asking the Power BI Community
Share access to a semantic model
Article • 02/14/2024

To make it possible for other users to take advantage of a semantic model, you can
share it with them. Sharing a semantic model means granting access to it. This
document shows you how to grant access to a semantic model using the Share
semantic model dialog.

Share a semantic model

To share a semantic model

1. From either the semantic model's options menu on the OneLake data hub or from
the data details page, choose Share as follows:

OneLake data hub: In the data items list, open the options menu and select
Share. On a recommended data item tile, choose Share on the More options
(…) menu.

Semantic model details page: Click the Share icon on the action bar at the
top of the page.
2. In the Share semantic model dialog that appears, enter the names or email
addresses of the specific people or groups (distribution groups or security groups)
that you want to grant access to, then choose the types of access you wish to
grant. You can optionally choose to send them an email notifying them that
they've been granted access.

Allow allow recipients to modify this semantic model: This option allows the
recipients to modify the semantic model.
 Allow recipients to share this semantic model: This option allows the
recipients to grant access to other users via sharing.

Allow recipients to build content with the data associated with this
semantic model: This option grants the recipients Build permission on the
semantic model, which enables them to build new reports and dashboards
based on the data associated it.

If you clear this checkbox, the user will get read-only permission on the
semantic model. Read-only permission allows them to explore the semantic
model on the semantic model's info page but doesn't allow them to build
new content based on the semantic model.

Send an email notification: When this option is selected, an email will be

sent to the recipients notifying them that they have been granted access to
the semantic model. You can add an optional message to the email message.

3. Click Grant access.

７ Note

When you press Grant access, access is granted automatically. No further approval
is required.

To monitor, change, or remove user access to your semantic model, see Manage
semantic model access permissions.

Related content
Semantic model permissions
Manage semantic model access permissions
Use semantic models across workspaces
Share a report via link
Questions? Try asking the Power BI Community
Semantic model permissions
Article • 04/05/2024

This article describes semantic model permissions in the Power BI service and how these
permissions are acquired by users.

What are the semantic model permissions?

The table below describes the four levels of permission that control access to semantic
models in the Power BI service. It also describes the permissions that the semantic
model owner has on the semantic model, and other actions that only the semantic
model owner can perform.

ﾉ Expand table

Permission Description

Read Allows user to access reports and other solutions, such as composite models on
Premium/PPU workspaces, that read data from the semantic model.
Allows user to view semantic model settings.

Build Allows user to build new content from the semantic model, as well as find content
that uses the semantic model.
Allows user to access reports that access composite models on Power BI Pro
workspaces.
Allows user to build composite models.
Allows user to pull the data into Analyze in Excel.
Allows querying using external APIs such as XMLA.
Allows user to see hidden data fields.

Reshare Allows user to grant semantic model access.

Write Allows user to republish the semantic model.

Allows user to backup and restore the semantic model.
Allows user to make changes to the semantic model via XMLA.
Allows user to edit semantic model settings, except data refresh, credentials, and
automatic aggregations.

Owner The semantic model owner is not a permission per se, but rather a conceptual role
that has all the permissions on a semantic model. The first semantic model owner is
the person who created the semantic model, and afterwards the last person to
configure the semantic model after taking it over in the semantic model settings.

In addition to the permissions above that can be granted explicitly, a semantic

 Permission Description

model owner can configure semantic model refresh, credentials, and automatic
aggregations.

How are the semantic model permissions

acquired?

Permissions acquired implicitly via workspace role

A user's role in a workspace implicitly grants them permissions on the semantic models
in the workspace, as described in the following table.

ﾉ Expand table

Admin Member Contributor Viewer

Read

Build

Reshare

Write

７ Note

Permissions inherited via workspace role can only be changed or taken away from a
user by changing or removing their role in the workspace. They can't be changed or
removed explicitly using the manage permissions page.

Permissions granted explicitly via the manage semantic

model permissions page
A user with an Admin or Member role in the workspace can explicitly grant permissions
to other users using the manage permissions page.

Permissions acquired via a link

When users share reports or semantic models, links are created that provide permissions
on the semantic model. Users authorized to use those links will be able to access the
semantic model. Users with Admin or Member roles in the workspace where a semantic
model is located can manage these links on the manage permissions page.

Permissions granted in an app

Users may acquire permissions on a semantic model used in an app if the app owner
allows this in the app permissions configuration.

Permissions granted via REST APIs

Semantic model permissions can be set via REST APIs. For more information, see
Semantic model permissions in the context of the Power BI REST APIs.

Semantic model permissions and row-level

security (RLS)
Row-level security may affect the ability of users with read or build permission on a
semantic model to read data from the semantic model.

When RLS isn't defined on the semantic model, users with write, read, or build
permission on the semantic model can read data from the semantic model.
When RLS is defined on the semantic model:
Users with only read or build permission on the semantic model will not be able
to read data from the semantic model unless they belong to one of its RLS
roles.
Users with write permission on the semantic model will be able to read data
from the semantic model regardless of whether or not they belong to any of its
RLS roles.

Related content
Share access to a semantic model
Manage semantic model permissions
Semantic model permissions in the context of the Power BI REST APIs

Feedback
Was this page helpful?
  Yes  No

Provide product feedback | Ask the community

Manage semantic model access
permissions (preview)
Article • 11/10/2023

The semantic model manage permissions page enables you to monitor and manage
access to your semantic model. It has two tabs that help you control access to your
semantic model:

Direct access: Enables you to monitor, add, modify, or delete access permissions
for specific people or groups (distribution groups or security groups).
Shared report links: Shows you links that were generated for sharing reports. Such
links sometimes also give access to your semantic model. On this tab you can
review them and remove them if necessary.

This document explains how to use the semantic model manage permissions page.

７ Note

In order to be able to access a semantic model's manage permissions page, you

must have an admin or member role in the workspace where the semantic model
is located.

Open the semantic model manage permissions

page
To open the semantic model manage permissions page:

From the OneLake data hub: Select Manage permissions on the More options (…)
menu.
From the semantic model details page: Select the Share icon on the action bar at
the top of the page and choose Manage permissions.

From the Share semantic model dialog: In the dialog header, select Manage
permissions on the More options (…) menu. This opens the Manage permissions
side pane. In the side pane, choose Advanced at the bottom of the pane.
These actions will open the semantic models manage permissions page. The manage
permissions page has two tabs to help you manage semantic model access.

Manage direct access

The direct access tab lists users who have been granted access. For each user, you can
see their email address and the permissions they have.

To modify a user’s permissions, select More options (…) and choose one of the
available options.
 To grant semantic model access to another user, click + Add user. The Share
semantic model dialog will open.

Managing permissions granted through an app

Permissions on the semantic model that have been granted through an app are
indicated by the word "App" followed by the permissions enclosed in parentheses, as
shown in the image below.

You can't modify permissions granted through an app directly from the Direct access tab
- you must first remove them from the app configuration. To remove such permissions:

1. Edit the app and unselect the relevant permissions on the Permissions tab of the
app's configuration settings.

2. Republish the app.

3. Go to the Direct access tab of the semantic model's manage semantic model
permissions page as described above. The user will still have the permissions that
were granted via the app before update, but now they won't be tied to the app
 (note that the parentheses are gone). Now you can remove whatever permissions
you desire.

Manage links generated for report sharing

The shared report links tab lists links that have been created to shared reports that are
based on your semantic model. Such links may also grant access to the report’s
underlying semantic model, and so they are listed here. You can see what permissions
the link carries and who created the link. You can also delete the link from the system if
you so desire.

２ Warning

Deleting a link removes it from the system. Users who use the link to access a
report may lose access to that report.

Next steps
Semantic model permissions
Share access to a semantic model
Use semantic models across workspaces
Share a report via link
Questions? Try asking the Power BI Community
Build permission for shared semantic
models
Article • 11/10/2023

When you create a report in Power BI Desktop, the data in that report is stored in a data
model. When you publish a report to the Power BI service, the data model is also
published to the service as a semantic model at the same time. When you share the
report with others, you can give them Build permission for the semantic model that the
report is built on, so they can discover and reuse it for their own reports, dashboards,
etc. This article explains how you control access to the semantic model using Build
permission.

Build permission applies to semantic models. When you give users Build permission,
they can build new content on your semantic model, such as reports, dashboards,
pinned tiles from Q&A, paginated reports, and Insights discovery. If a report outside the
semantic model workspace uses your semantic model, you can't delete the semantic
model. If you try to do so, you get an error message.

Users also need Build permission to do the following actions:

Export underlying Power BI data.

Build new content on the semantic model, such as with Analyze in Excel.
Access the data via the XML for Analysis (XMLA) endpoint.

How users get Build permission

Users get Build permission for a semantic model in a few different ways:

Users that have at least a Contributor role in a workspace have Build permission on
the semantic models in that workspace, as well as permission to copy reports in
that workspace. For more information about roles in workspaces, see Roles in
workspaces in Power BI.

Semantic model owners can assign Build permission to specific users or security
groups on the Manage permissions page. For more information, see Manage
semantic model access permissions.

A user with an Admin or Member role in the workspace where the semantic model
resides can decide during app publishing that users with permission for the app
also get Build permission for the underlying semantic models. For more
information, see Create and manage multiple audiences.
 If you have Reshare and Build permission on a semantic model, and you share a
report or dashboard you built on that semantic model, you can specify that the
recipients also get Build permission for the semantic model. For more information,
see Share Power BI reports and dashboards with coworkers and others.

Remove Build permission

To remove Build permission for users of a shared semantic model, follow the instructions
at Manage direct access.

If you remove Build permission, the people whose permission you revoked can still see
the report, but can no longer edit the report or export underlying data. Users with only
read permission can still export summarized data.

Remove Build permission for a semantic model in an app

If you distribute an app from a workspace, removing people's access to the app doesn't
automatically remove their build and reshare permissions. To remove their Build
permissions, take the following steps:

1. In the workspace, in list view, select Update app.

2. Select the Audience tab, and then in the Manage Audience Access side pane,
hover over the person or group whose access you want to delete and select the
trash icon that appears. When you're done, select Update app.
 You'll see a message that you need to go to Manage permissions to remove
permissions for users with existing access.

3. Select Update.
 4. Follow the instructions at Manage permissions to see how to remove permissions
from users with existing access. When you take away a user's Build permission on a
semantic model, they can still see reports built on the semantic model, but they
can no longer edit the reports.

Configure how users request Build permission

Certain actions, such as creating a report based on a semantic model or accessing the
details page of a semantic model in the data hub, require Build permission on the
semantic model. By default, when users who don't have Build permission try these
actions, they get a dialog box that lets them send email to the semantic model owner
requesting Build permission. The email includes the user's details, the name of the
semantic model they’re requesting access to, and any other information they optionally
provide.

Change the access request behavior

If you have an Admin, Member, or Contributor role in the workspace where the semantic
model resides, you can change the default access request behavior for a semantic model
by going to the semantic model's settings and configuring the Request access options
as desired.
The default option, not selected in the preceding image, is for Build permission
requests to come to you via email. You're responsible for acting on the requests
and notifying the requestors.

The second option is for you to provide instructions about how to get Build
permission, rather than receiving requests via email. You might choose this option,
for example, if your organization uses an automated system for handling access
requests. When users who don't have Build permission try an action that requires
Build permission, they see a message with the instructions you provide.

The Instructions text area in the preceding Request access example shows sample
instructions. Instructions must be in plain text. HTML or any other type of code
formatting render as plain text, rather than the code format. The following example
shows the instructions users see when they try an action they need Build
permission for.
 ７ Note

When you provide specific instructions, your email address is visible to users
who request access.

More granular permissions

Power BI provides Build permission as a complement to Read and Reshare permissions.
All users who already have Read permission for semantic models via app permissions,
sharing, or workspace access also get Build permission for those semantic models.
Those users get Build permission automatically because Read permission already grants
them the right to build new content on the semantic model by using Analyze in Excel or
Export.

With the more granular Build permission, you can choose who can only view the content
in an existing report or dashboard, and who can create content connected to the
underlying semantic model.

Next steps
Use semantic models across workspaces
Share a semantic model
Roles in workspaces
Manage semantic model access permissions

Questions? Try asking the Power BI Community

Create reports based on semantic
models from different workspaces
Article • 11/10/2023

Learn how you can create reports in your own workspaces based on semantic models in
other workspaces. To build a report on top of an existing semantic model, you can start
from Power BI Desktop or from the Power BI service, in your My workspace or in
another workspace.

In the Power BI service: Create > Report > Pick a published semantic model.

In Power BI Desktop: from the Home ribbon, select Get data > Power BI semantic
models.

![Screenshot shows how to connect to an existing semantic model in the Power BI

service and Power BI Desktop.](media/service-semantic models-across-
workspaces/power-bi-connect-semantic model-pk.png)

In both cases, the semantic model discovery experience starts in the Data hub. You see
all the semantic models that you have access to, regardless of where they are:

One of the semantic models is labeled Promoted. Learn about that label in [Find an
endorsed semantic model](#find-an-endorsed-semantic model), later in this article.

The semantic models in this list meet at least one of the following conditions:
 The semantic model is in a workspace that you're a member of. See
[Considerations and limitations](service-semantic models-across-
workspaces.md#considerations-and-limitations).
You have Build permission for the semantic model.
The semantic model is in your My workspace.

７ Note

If you're a free user, you see only semantic models in your My workspace, or
semantic models for which you have Build permission that are in Premium-capacity
workspaces.

When you select Create, you create a live connection to the semantic model. The report
creation experience opens with the full semantic model available. You haven't made a
copy of the semantic model. The semantic model still resides in its original location. You
can use all tables and measures in the semantic model to build your own reports. Row-
level security (RLS) restrictions on the semantic model are in effect, so you only see data
you have permissions to see based on your RLS role.

You can save the report to the current workspace in the Power BI service, or publish the
report to a workspace from Power BI Desktop. Power BI automatically creates an entry in
the list of semantic models if the report is based on a semantic model outside of the
workspace.

The entry shows information about the semantic model, and a few select actions.

Find an endorsed semantic model

There are two different kinds of endorsed semantic models. Semantic model owners can
promote a semantic model that they recommend to you. Also, the Power BI admin can
designate experts in your organization who can certify semantic models for everyone to
use. Promoted and certified semantic models both display badges that you see both
when looking for a semantic model, and in the list of semantic models in a workspace.
The name of the person who certified a semantic model is displayed in a tooltip during
the semantic model discovery experience. Hover over the Certified label and you see it.

In the Power BI service: OneLake data hub.

 In Power BI Desktop: Get data > Power BI semantic models.

In OneLake data hub, select Endorsed in your org.

Next steps
Use semantic models across workspaces
Questions? Try asking the Power BI Community
Copy reports from other workspaces
Article • 11/10/2023

When you find a report you like in a workspace or an app, you can make a copy of it
and save it to a different workspace. Then you can modify your copy of the report,
adding or deleting visuals and other elements. You don't have to worry about creating
the data model - the copy of report will still reference the same semantic model as the
original report. And it's much easier to modify an existing report than it is to start from
scratch. However, when you make an app from your workspace, sometimes you can't
publish your copy of the report in the app. See Considerations and limitations in the
article "Use semantic models across workspaces" for details.

Prerequisites
To copy a report, you need a Pro or Premium Per User (PPU) license, even if the
original report is in a workspace in a Premium capacity.
To copy a report to another workspace, or to create a report in one workspace
based on a semantic model in another workspace, you need Build permission for
the semantic model. If you have at least the Contributor role in the workspace
where the semantic model is located, you automatically have Build permission
through your workspace role. You also need at least the Contributor role in the
workspace where the report you're copying is located, and in the workspace where
you want to create the copy of the report. See Roles in workspaces for details.

Save a copy of a report in a workspace

1. In a workspace, find a report in the list. Open the More options menu and select
Save a copy.
 You only see the Save a copy option if you have Build permission. Even if you have
access to the workspace, you have to have Build permission for the semantic
model.

2. In Save a copy of this report, give the report a name and select the destination
workspace.
 You can save the report to the current workspace or a different one in the Power BI
service. You only see workspaces in which you're a member.

3. Select Save.

Power BI automatically creates a copy of the report in the workspace you selected.
In the list view of that workspace, you won't see the referenced semantic model if
it is located in another workspace. To see the shared semantic model, on the report
copy in list view select More options > View lineage.


 In lineage view, semantic models that are located in other workspaces show the
name of the workspace they're located in. This makes it easy to see which reports
and dashboards use semantic models that are outside the workspace.

See Your copy of the report in this article for more about the report and related
semantic model.

Copy a report in an app

1. In an app, open the report you want to copy.

2. In the menu bar, select File > Save a copy.

You only see the Save a copy option if app permissions grant Build permission for the
underlying semantic model, and allow users to make copies of the report.

3. Give your report a name, select a destination workspace, and then select Save.

Your copy is automatically saved to the workspace you selected.

 4. Select Go to report to open your copy.

Your copy of the report

When you save a copy of the report, you create a live connection to the semantic model,
and you can open the report creation experience with the full semantic model available.

You haven't made a copy of the semantic model. The semantic model still resides in its
original location. You can use all tables and measures in the semantic model in your own
report. Row-level security (RLS) restrictions on the semantic model are in effect, so you
only see data you have permissions to see based on your RLS role.

View related semantic models

When you have a report in one workspace based on a semantic model in another
workspace, you may need to know more about the semantic model it's based on.

1. In the report, select More options > See related content.

2. The Related content dialog box shows all related items. In this list, the semantic
model looks like any other. There is no indication of where the semantic model
resides.
Delete a report copy
If you want to delete the copy of the report, in the list of reports in the workspace, hover
over the report you want to delete, select More options, and choose Delete.
 ７ Note

Deleting a report doesn't delete the semantic model it is built on.

Next steps
Use semantic models across workspaces
Questions? Try asking the Power BI Community
Control the use of semantic models
across workspaces
Article • 04/02/2024

Using semantic models across workspaces is a powerful way to drive data culture and
data democratization within an organization. Still, if you're a Power BI admin, sometimes
you want to restrict the flow of information within your Power BI tenant. With the tenant
setting Use semantic models across workspaces, you can restrict semantic model reuse
either completely or partially per security groups.

Some of the effects of turning off this setting are listed below:

The button to copy reports across workspaces isn't available.

In a report based on a shared semantic model, the Edit report button isn't
available.
In the Power BI service, the discovery experience only shows semantic models in
the current workspace.
In Power BI Desktop, the discovery experience only shows semantic models from
workspaces where you're a member.
In the Data hub, users see semantic models that were shared with them outside of
the workspace, but they can't interact with them.
In Power BI Desktop, if users open a .pbix file with a live connection to a semantic
model outside any workspaces they are a member of, they see an error message
 asking them to connect to a different semantic model. See the "Limitations"
section of Download a report from the Power BI service to Power BI Desktop for
more information.

Provide a link for the certification process

As a Power BI admin, you can provide a URL for the Learn more link on the
Endorsement setting page. See Enable content certification for detail. This link can go to
documentation about your certification process. If you don't provide a destination for
the Learn more link, by default it points to the Endorse your content article.

Related content
Use semantic models across workspaces
Questions? Try asking the Power BI Community

Feedback
Was this page helpful?  Yes  No
Provide product feedback | Ask the community
Azure and Power BI
Article • 01/12/2023

With Azure services and Power BI, you can turn your data processing efforts into
analytics and reports that provide real-time insights into your business. Whether your
data processing is cloud-based or on-premises, straightforward, or complex, single-
sourced or massively scaled, warehoused, or real-time, Azure and Power BI have the
built-in connectivity and integration to bring your business intelligence efforts to life.

Power BI has a multitude of Azure connections available, and the business intelligence
solutions you can create with those services are as unique as your business. You can
connect as few as one Azure data source, or a handful, then shape and refine your data
to build customized reports.

Azure SQL Database and Power BI

You can start with a straightforward connection to an Azure SQL Database, and create
reports to monitor the progress of your business. Using the Power BI Desktop, you can
create reports that identify trends and key performance indicators that move your
business forward.
There's plenty more information for you to learn about Azure SQL Database .

Transform, shape, and merge your cloud data

Do you have more complex data, and all sorts of sources? No problem. With Power BI
Desktop and Azure services, connections are just a tap of the Get Data dialog away.
Within the same query you can connect to your Azure SQL Database, your Azure
HDInsight data source, and your Azure Blob Storage or Azure Table Storage. Then select
only the subsets within each that you need, and refine it from there.

You can create different reports for different audiences too, using the same data
connections and even the same query. Just build a new report page, refine your
visualizations for each audience, and watch it keep the business in the know.
For more information, take a look at the following resources:

Azure SQL Database

Azure HDInsight
Azure Storage (Blob storage and Table storage)

And for more information about Azure resources available in Power BI, see Azure data
sources.

Get complex (and ahead) using Azure services

and Power BI
You can expand as much as you need with Azure and Power BI. Harness multi-source
data processing, make use of massive real-time systems, use Stream Analytics and
Event Hubs , and coalesce your varied SaaS services into business intelligence reports
that give your business an edge.
Context insights with Power BI Embedded
analytics
Embed stunning, interactive data visualizations in applications, websites, portals, and
more, to take advantage of your business data. With Power BI Embedded as a resource
in Azure , you can easily embed interactive reports and dashboards, so your users can
enjoy consistent, high-fidelity experiences across devices. Power BI used with
embedding analytics is to help you through your journey from Data to Knowledge to
Insights to Actions. Furthermore, you can extend the value of Power BI and Azure also
by embedding analytics in your organization's internal applications and portals .

There's lots of information about Power BI APIs in the Power BI Developer Portal .

For more information, see Power BI Embedded.

Embed your Power BI data within your app

Embed stunning, interactive data visualizations in applications, websites, portals, and
more, to showcase your business data in context. Using Power BI Embedded in Azure ,
you can easily embed interactive reports and dashboards, so your users can enjoy
consistent, high-fidelity experiences across devices.

What could you do with Azure and Power BI?

There are all sorts of scenarios where Azure and Power BI can be combined. The
possibilities and opportunities are as unique as your business. For more information
about Azure services, check out this overview page, which describes Data Analytics
Scenarios using Azure, and learn how to transform your data sources into intelligence
that drives your business ahead.
Power BI and Azure egress
Article • 01/23/2024

Data moving out, or egress, of Azure data centers can incur bandwidth charges. When
using Power BI with Azure data sources, you can avoid Azure egress charges by making
sure your Power BI service tenant is in the same region as your Azure data sources.

When your Power BI service tenant is deployed in the same Azure region as you deploy
your data sources, you don't incur egress charges for scheduled refresh and DirectQuery
interactions.

Determining where your Power BI tenant is

located
To find out where your Power BI tenant is located, see Find the default region for your
organization.

For Power BI Premium Multi-Geo customers, if your Power BI tenant isn't in the optimal
location for some of your Azure-based data sources, you can deploy Power BI Premium
Multi-Geo in the desired Azure region and benefit from having your Power BI tenant
and Azure data sources in the same Azure region.

７ Note

Power BI Premium Per User (PPU) is not supported for Multi-Geo.

Related content
For more information about Power BI Premium or Multi-Geo, take a look at the
following resources:

Azure bandwidth pricing details

What is Microsoft Power BI Premium?
How to purchase Power BI Premium
Multi-Geo support for Power BI Premium
Where is my Power BI tenant located?
Power BI Premium FAQ
Azure Synapse Analytics (formerly SQL
Data Warehouse) with DirectQuery
Article • 11/10/2023

Azure Synapse Analytics (formerly SQL Data Warehouse) with DirectQuery allows you to
create dynamic reports based on data and metrics you already have in Azure Synapse
Analytics. With DirectQuery, queries are sent back to your Azure Synapse Analytics in
real time as you explore the data. Real-time queries, combined with the scale of Synapse
Analytics enables users to create dynamic reports in minutes against terabytes of data.

When you use the Azure Synapse Analytics connector:

Specify the fully qualified server name when you connect (see details later in this
article).
Ensure firewall rules for the server are configured to "Allow access to Azure
services".
Every action such as selecting a column or adding a filter will directly query the
data warehouse.
Tiles are set to refresh approximately every 15 minutes and you don't need to
schedule a refresh. You can adjust refresh in the Advanced settings when you
connect.
Q&A isn't available for DirectQuery semantic models.
Schema changes aren't picked up automatically.

These restrictions and notes can change as we continue to improve the experience.
Steps to connect are in the next section.

Build dashboards and reports in Power BI

） Important

We continually improve connectivity to Azure Synapse Analytics. For the best

experience to connect to your Azure Synapse Analytics data source, use Power BI
Desktop. After you've built your model and report, you can publish it to the Power
BI service. The previously available direct connector for Azure Synapse Analytics in
the Power BI service is no longer available.

The easiest way to move between your Synapse Analytics and Power BI is to create
reports in Power BI Desktop. To get started, download and install Power BI Desktop.
Connect through Power BI Desktop
You can connect to an Azure Synapse Analytics using the process described in the Power
Query article about Azure SQL Data Warehouse.

Find Parameter Values

Your fully qualified server name and database name can be found in the Azure portal.
Azure Synapse Analytics only has a presence in the Azure portal at this time.

７ Note

If your Power BI tenant is in the same region as the Azure Synapse Analytics there
will be no egress charges. To find where your Power BI tenant is located, see Find
the default region for your organization.

Single sign-on
After you publish an Azure SQL DirectQuery semantic model to the service, you can
enable single sign-on (SSO) using Microsoft Entra ID OAuth2 for your end users.

To enable SSO, go to settings for the semantic model, open the Data Sources tab, and
check the SSO box.
When the SSO option is enabled and your users access reports built atop the data
source, Power BI sends their authenticated Microsoft Entra credentials in the queries to
the Azure SQL database or data warehouse. This option enables Power BI to respect the
security settings that are configured at the data source level.

The SSO option takes affect across all semantic models that use this data source. It does
not affect the authentication method used for import scenarios.

７ Note

For SSO to work properly, the semantic model must be on the same tenant as the
Azure SQL resource.

Related content
DirectQuery in Power BI
What is Power BI?
Data sources for the Power BI service
What is dedicated SQL pool (formerly SQL DW) in Azure Synapse Analytics?

More questions? Ask the Power BI Community

Azure SQL Database with DirectQuery
Article • 01/06/2023

Learn how you can connect directly to Azure SQL Database and create reports that use
live data. You can keep your data at the source and not in Power BI.

With DirectQuery, queries are sent back to your Azure SQL Database as you explore the
data in the report view. This experience is suggested for users who are familiar with the
databases and entities they connect to.

） Important

This description assumes that Azure SQL database is not behind a VNET or has
private link endpoint enabled.

Notes:

Specify the fully qualified server name when connecting (see below for more
details).
Ensure firewall rules for the database are configured to"Allow access to Azure
services.
Every action such as selecting a column or adding a filter will send a query back to
the database.
Tiles are refreshed every hour (refresh doesn't need to be scheduled). You can
adjust how often to refresh in the Advanced settings when you connect.
Schema changes aren't picked up automatically.
Changing the data source connection string alias from xxxx.database.windows.net
to xxxx.domain.com indicates to the Power BI service that it's an on-premises
datasource and always requires a gateway connection to be established.

These restrictions and notes may change as we continue to improve the experiences.
The steps to connect are detailed below.

） Important

We have been improving our connectivity to Azure SQL Database. For the best
experience to connect to your Azure SQL Database data source, use Power BI
Desktop. Once you've built your model and report, you can publish it to the Power
BI service. The direct connector for Azure SQL Database in the Power BI service is
now deprecated.
Power BI Desktop and DirectQuery
To connect to Azure SQL Database using DirectQuery, you must use Power BI Desktop.
This approach provides more flexibility and capabilities. Reports created using Power BI
Desktop can then be published to the Power BI service. To learn more about how to
connect to Azure SQL Database in Power BI Desktop, see Use DirectQuery in Power BI
Desktop.

Find parameter values

You can find your fully qualified server name and database name in the Azure portal.

Single sign-on
After you publish an Azure SQL DirectQuery dataset to the service, you can enable
single sign-on (SSO) using Azure Active Directory (Azure AD) OAuth2 for your end users.

To enable SSO, go to settings for the dataset, open the Data Sources tab, and check the
SSO box.
When the SSO option is enabled and your users access reports built atop the data
source, Power BI sends their authenticated Azure AD credentials in the queries to the
Azure SQL database or data warehouse. This option enables Power BI to respect the
security settings that are configured at the data source level.

The SSO option takes affect across all datasets that use this data source. It does not
affect the authentication method used for import scenarios.

７ Note

For SSO to work properly, the dataset must be on the same tenant as the Azure
SQL resource.

Next steps
Use DirectQuery in Power BI Desktop
What is Power BI?
Data sources for the Power BI service

More questions? Try the Power BI community

Data refresh in Power BI
Article • 05/03/2024

Power BI enables you to go from data to insight to action quickly, yet you must make
sure the data in your Power BI reports and dashboards is recent. Knowing how to refresh
the data is often critical in delivering accurate results.

This article describes the data refresh features of Power BI and their dependencies at a
conceptual level. It also provides best practices and tips to avoid common refresh issues.
The content lays a foundation to help you understand how data refresh works. For
targeted step-by-step instructions to configure data refresh, refer to the tutorials and
how-to guides listed in the Related content section at the end of this article.

Understanding data refresh

Embed Power BI content with service principal and an application secret

Whenever you refresh data, Power BI must query the underlying data sources, possibly
load the source data into a semantic model, and then update any visualizations in your
reports or dashboards that rely on the updated semantic model. The entire process
consists of multiple phases, depending on the storage modes of your semantic models,
as explained in the following sections.

To understand how Power BI refreshes your semantic models, reports, and dashboards,
you must be aware of the following concepts:

Storage modes and semantic model types: The storage modes and semantic
model types that Power BI supports have different refresh requirements. You can
choose between reimporting data into Power BI to see any changes that occurred
or querying the data directly at the source.
Power BI refresh types: Regardless of semantic model specifics, knowing the
various refresh types can help you understand where Power BI might spend its
time during a refresh operation. And combining these details with storage mode
specifics helps to understand what exactly Power BI performs when you select
Refresh now for a semantic model.

Storage modes and semantic model types

A Power BI semantic model can operate in one of the following modes to access data
from various data sources. For more information, see Storage mode in Power BI
Desktop.
 Import mode
DirectQuery mode
LiveConnect mode
Push mode

The following diagram illustrates the different data flows, based on storage mode. The
most significant point is that only Import mode semantic models require a source data
refresh. They require refresh because only this type of semantic model imports data
from its data sources, and the imported data might be updated on a regular or ad-hoc
basis. DirectQuery semantic models and semantic models in LiveConnect mode to
Analysis Services don't import data; they query the underlying data source with every
user interaction. Semantic models in push mode don't access any data sources directly
but expect you to push the data into Power BI. Semantic model refresh requirements
vary depending on the storage mode/semantic model type.

Semantic models in Import mode

Power BI imports the data from the original data sources into the semantic model.
Power BI report and dashboard queries submitted to the semantic model return results
from the imported tables and columns. You might consider such a semantic model a
point-in-time copy. Because Power BI copies the data, you must refresh the semantic
model to fetch changes from the underlying data sources.

When a semantic model is refreshed, it's either fully refreshed or partially refreshed.
Partial refresh will take place in semantic models that have tables with an incremental
refresh policy. In these semantic models, only a subset of the table partitions are
refreshed. In addition, advanced users can use the XMLA endpoint to refresh specific
partitions in any semantic model.

The amount of memory required to refresh a semantic model depends on whether

you're performing a full or partial refresh. During the refresh, a copy of the semantic
model is kept to handle queries to the semantic model. This means that if you're
performing a full refresh, you'll need twice the amount of memory the semantic model
requires.

We recommend that you plan your capacity usage to ensure that the extra memory
needed for semantic model refresh, is accounted for. Having enough memory prevents
refresh issues that can occur if your semantic models require more memory than
available, during refresh operations. To find out how much memory is available for each
semantic model on a Premium capacity, refer to the Capacities and SKUs table.

For more information about large semantic models in Premium capacities, see large
semantic models.

Semantic models in DirectQuery mode

Power BI doesn't import data over connections that operate in DirectQuery mode.
Instead, the semantic model returns results from the underlying data source whenever a
report or dashboard queries the semantic model. Power BI transforms and forwards the
queries to the data source.

７ Note

Live connection reports submit queries to the capacity or Analysis Services instance
that hosts the semantic model or the model. When using external analysis services
such as SQL Server Analysis Services (SSAS) or Azure Analysis Services (AAS),
resources are consumed outside of Power BI.

Because Power BI doesn't import the data, you don't need to run a data refresh.
However, Power BI still performs tile refreshes and possibly report refreshes, as the next
section on refresh types explains. A tile is a report visual pinned to a dashboard, and
dashboard tile refreshes happen about every hour so that the tiles show recent results.
You can change the schedule in the semantic model settings, as in the screenshot below,
or force a dashboard update manually by using the Refresh now option.

７ Note

Semantic models in import mode and composite semantic models that

combine import mode and DirectQuery mode don't require a separate tile
refresh, because Power BI refreshes the tiles automatically during each
scheduled or on-demand data refresh. Semantic models that are updated
based on the XMLA endpoint will only clear the cached tile data (invalidate
cache). The tile caches aren't refreshed until each user accesses the
dashboard. For import models, you can find the refresh schedule in the
"Scheduled refresh" section of the Semantic models tab. For composite
semantic models, the "Scheduled refresh" section is located in the Optimize
Performance section.
Power BI does not support cross-border live connections to Azure Analysis
Services (AAS) in a sovereign cloud.
Push semantic models
Push semantic models don't contain a formal definition of a data source, so they don't
require you to perform a data refresh in Power BI. You refresh them by pushing your
data into the semantic model through an external service or process, such as Azure
Stream Analytics. This is a common approach for real-time analytics with Power BI.
Power BI still performs cache refreshes for any tiles used on top of a push semantic
model. For a detailed walkthrough, see Tutorial: Stream Analytics and Power BI: A real-
time analytics dashboard for streaming data.

Power BI refresh types

A Power BI refresh operation can consist of multiple refresh types, including data
refresh, OneDrive refresh, refresh of query caches, tile refresh, and refresh of report
visuals. While Power BI determines the required refresh steps for a given semantic
model automatically, you should know how they contribute to the complexity and
duration of a refresh operation. For a quick reference, refer to the following table.

ﾉ Expand table

Storage Data refresh OneDrive Query caches Tile refresh Report

mode refresh visuals

Import Scheduled Yes, for If enabled on Automatically No

and on- connected Premium and on-demand
demand semantic capacity
models

DirectQuery Not Yes, for Not applicable Automatically No

applicable connected and on-demand
semantic
models

LiveConnect Not Yes, for Not applicable Automatically Yes

applicable connected and on-demand
semantic
models

Push Not Not applicable Not practical Automatically No

applicable and on-demand

Another way to consider the different refresh types is what they impact and where you
can apply them. Changes in data source table structure, or schema, such as a new,
renamed, or removed column can only be applied in Power BI Desktop, and in the
Power BI service they can cause the refresh to fail. For a quick reference on what they
impact, refer to the following table.

ﾉ Expand table

Refresh of report visuals Data refresh Schema refresh

What do the Queries used to populate Data is refreshed from Any data source table
different visuals are refreshed. the data source. structure change since
refresh types previous refresh will
do? For visuals using Doesn't apply to show.
DirectQuery tables the visual DirectQuery tables as
will query to get the latest they are at the visual For example: To show a
data from the data source. level and rely on refresh new column added to
of report visuals. a Power BI Dataflow or
For visuals using imported SQL Database view.
tables the visual will only For imported tables the
query data already imported data is refreshed from Applies to both
to the semantic model on the source. imported and
the last data refresh. DirectQuery tables.

In Power BI Desktop refresh of report visuals, data refresh, and schema refresh all
happen together using

Home ribbon > Refresh button

Home ribbon > Transform data > Close & Apply button
The context menu (right-click or select the ellipsis) on any table then choosing
Refresh data

These refresh types cannot always be applied independently, and where you can apply
them is different in Power BI Desktop and the Power BI service. For a quick reference,
refer to the following table.

ﾉ Expand table

Refresh of report visuals Data refresh Schema refresh

In Power View ribbon > Not available independently Not available

BI Performance from other refresh types independently from
Desktop Analyzer button > other refresh types
Refresh visuals
Creating and
changing visuals
causing a DAX
query to run
When Page Refresh
is turned on
 Refresh of report visuals Data refresh Schema refresh

(DirectQuery only)
Opening the PBIX
file

In the When the browser Scheduled refresh Not available

Power BI loads or reloads the Refresh now
service report Refresh a Power BI
Clicking the Refresh semantic model from
Visuals top right Power Automate
menu bar button Processing the table
Clicking the Refresh from SQL Server
button in edit mode Management Studio
When Page Refresh (Premium)
is turned on
(DirectQuery only)

Keep in For example, if you open a Data refresh on the Power BI A renamed or removed
mind report in the browser, service will fail when the column or table at the
then the scheduled source column or table is data source will be
refresh performs a data renamed or removed. It fails updated with a schema
refresh of the imported because the Power BI service refresh in Power BI
tables, the report visuals doesn't also include a schema Desktop, but it can
in the open browser won't refresh. To correct this error, a break visuals and DAX
update until a refresh of schema refresh needs to expressions (measures,
report visuals is initiated. happen in Power BI Desktop calculated columns, row
and the semantic model level security, etc.), as
republished to the service. well as remove
relationships, that are
dependent on those
columns or tables.

Data refresh
For Power BI users, refreshing data typically means importing data from the original data
sources into a semantic model, either based on a refresh schedule or on-demand. You
can perform multiple semantic model refreshes daily, which might be necessary if the
underlying source data changes frequently. Power BI limits semantic models on shared
capacity to eight scheduled daily semantic model refreshes. The eight time values are
stored in the backend database and are based on the local time zone that was selected
on the Semantic model Settings page. The scheduler checks which model should be
refreshed and at what time(s). The quota of eight refreshes resets daily at 12:01 a.m.
local time.
If the semantic model resides on a Premium capacity, you can schedule up to 48
refreshes per day in the semantic model settings. For more information, see Configure
scheduled refresh later in this article. Semantic models on a Premium capacity with the
XMLA endpoint enabled for read-write support unlimited refresh operations when
configured programmatically with TMSL or PowerShell.

It's also important to call out that the shared-capacity limitation for daily refreshes
applies to both scheduled refreshes and API refreshes combined. You can also trigger an
on-demand refresh by selecting Refresh now in the semantic model menu, as the
following screenshot depicts. On-demand refreshes aren't included in the refresh
limitation. Also note that semantic models on a Premium capacity don't impose
limitations for API refreshes. If you're interested in building your own refresh solution by
using the Power BI REST API, see semantic models - Refresh semantic model.

７ Note

Data refreshes must complete in less than 2 hours on shared capacity. If your
semantic models require longer refresh operations, consider moving the semantic
model onto a Premium capacity. On Premium, the maximum refresh duration is 5
hours, but using XMLA endpoint to refresh data can bypass the 5-hour limit.

OneDrive refresh
If you created your semantic models and reports based on a Power BI Desktop file, Excel
workbook, or comma separated value (.csv) file on OneDrive or SharePoint Online,
Power BI performs another type of refresh, known as OneDrive refresh. For more
information, see Get data from files for Power BI.

Unlike a semantic model refresh during which Power BI imports data from a data source
into a semantic model, OneDrive refresh synchronizes semantic models and reports with
their source files. By default, Power BI checks about every hour if a semantic model
connected to a file on OneDrive or SharePoint Online requires synchronization.

Power BI performs refresh based on an item ID in OneDrive, so be thoughtful when

considering updates versus replacement. When you set a OneDrive file as the data
source, Power BI references the item ID of the file when it performs the refresh. Consider
the following scenario: you have a master file A and a production copy of that file B, and
you configure OneDrive refresh for file B. If you then copy file A over file B, the copy
operation deletes the old file B and creates a new file B with a different item ID, which
breaks OneDrive refresh. To avoid that situation, you can instead upload and replace file
B, which keeps its same item ID.

You can move the file to another location (using drag and drop, for example) and
refresh will continue to work because Power BI still knows the file ID. However, if you
copy that file to another location, a new instance of the file and a new fileID is created.
Therefore, your Power BI file reference is no longer valid and refresh will fail.

７ Note

It can take Power BI up to 60 minutes to refresh a semantic model, even once the
sync has completed on your local machine and after you've used Refresh now in the
Power BI service.

To review past synchronization cycles, check the OneDrive tab in the refresh history. The
following screenshot shows a completed synchronization cycle for a sample semantic
model.

As the above screenshot shows, Power BI identified this OneDrive refresh as a

Scheduled refresh, but it isn't possible to configure the refresh interval. You can only
deactivate OneDrive refresh in the semantic model's settings. Deactivating refresh is
useful if you don't want your semantic models and reports in Power BI to pick up any
changes from the source files automatically.
The semantic model settings page only shows the OneDrive Credentials and OneDrive
refresh sections if the semantic model is connected to a file in OneDrive or SharePoint
Online, as in the following screenshot. Semantic models that aren't connected to
sources file in OneDrive or SharePoint Online don't show these sections.

If you disable OneDrive refresh for a semantic model, you can still synchronize your
semantic model on-demand by selecting Refresh now in the semantic model menu. As
part of the on-demand refresh, Power BI checks if the source file on OneDrive or
SharePoint Online is newer than the semantic model in Power BI and synchronizes the
semantic model if so. The Refresh history lists these activities as on-demand refreshes
on the OneDrive tab.

Keep in mind that OneDrive refresh doesn't pull data from the original data sources.
OneDrive refresh simply updates the resources in Power BI with the metadata and data
from the .pbix, .xlsx, or .csv file, as the following diagram illustrates. To ensure that the
semantic model has the most recent data from the data sources, Power BI also triggers a
data refresh as part of an on-demand refresh. You can verify this in the Refresh history if
you switch to the Scheduled tab.
If you keep OneDrive refresh enabled for a OneDrive or SharePoint Online-connected
semantic model and you want to perform data refresh on a scheduled basis, make sure
you configure the schedule so that Power BI performs the data refresh after the
OneDrive refresh. For example, if you created your own service or process to update the
source file in OneDrive or SharePoint Online every night at 1 am, you could configure
scheduled refresh for 2:30 am to give Power BI enough time to complete the OneDrive
refresh before starting the data refresh.

Refresh of query caches

If your semantic model resides on a Premium capacity, you might be able to improve
the performance of any associated reports and dashboards by enabling query caching,
as in the following screenshot. Query caching instructs the Premium capacity to use its
local caching service to maintain query results, avoiding having the underlying data
source compute those results. For more information, see Query caching in Power BI
Premium.
Following a data refresh, however, previously cached query results are no longer valid.
Power BI discards these cached results and must rebuild them. For this reason, query
caching might not be as beneficial for reports and dashboards associated with semantic
models that you refresh often, for example 48 times per day.

Refresh of report visuals

This refresh process is less important because it's only relevant for live connections to
Analysis Services. For these connections, Power BI caches the last state of the report
visuals so that when you view the report again, Power BI doesn't have to query the
Analysis Services tabular model. When you interact with the report, such as by changing
a report filter, Power BI queries the tabular model and updates the report visuals
automatically. If you suspect that a report is showing stale data, you can also select the
Refresh button of the report to trigger a refresh of all report visuals, as the following
screenshot illustrates.
Only pinned visuals are refreshed, not pinned live pages. To refresh a pinned live page,
you can use the browser's Refresh button.

Review data infrastructure dependencies

Regardless of storage modes, no data refresh can succeed unless the underlying data
sources are accessible. There are three main data access scenarios:

A semantic model uses data sources that reside on-premises

A semantic model uses data sources in the cloud
A semantic model uses data from both, on-premises and cloud sources

Connecting to on-premises data sources

If your semantic model uses a data source that Power BI can't access over a direct
network connection, you must configure a gateway connection for this semantic model
before you can enable a refresh schedule or perform an on-demand data refresh. For
more information about data gateways and how they work, see What are on-premises
data gateways?

You have the following options:

Choose an enterprise data gateway with the required data source definition
Deploy a personal data gateway

７ Note
 You can find a list of data source types that require a data gateway in the article
Manage your data source - Import/Scheduled Refresh.

Using an enterprise data gateway

Microsoft recommends using an enterprise data gateway instead of a personal gateway

to connect a semantic model to an on-premises data source. Make sure the gateway is
properly configured, which means the gateway must have the latest updates and all
required data source definitions. A data source definition provides Power BI with the
connection information for a given source, including connection endpoints,
authentication mode, and credentials. For more information about managing data
sources on a gateway, see Manage your data source - import/scheduled refresh.

Connecting a semantic model to an enterprise gateway is relatively straightforward if

you're a gateway administrator. With admin permissions, you can promptly update the
gateway and add missing data sources, if necessary. In fact, you can add a missing data
source to your gateway straight from the semantic model settings page. Expand the
toggle button to view the data sources and select the Add to gateway link, as in the
following screenshot. If you aren't a gateway administrator, on the other hand, you must
contact a gateway admin to add the required data source definition.

７ Note

Only gateway admins can add data sources to a gateway. Also make sure your
gateway admin adds your user account to the list of users with permissions to use
the data source. The semantic model settings page only lets you select an
enterprise gateway with a matching data source that you have permission to use.
Make sure you map the correct data source definition to your data source. As the above
screenshot illustrates, gateway admins can create multiple definitions on a single
gateway connecting to the same data source, each with different credentials. In the
example shown, a semantic model owner in the Sales department would choose the
AdventureWorksProducts-Sales data source definition while a semantic model owner in
the Support department would map the semantic model to the
AdventureWorksProducts-Support data source definition. If the names of the data
source definition aren't intuitive, contact your gateway admin to clarify which definition
to pick.

７ Note

A semantic model can only use a single gateway connection. In other words, it is
not possible to access on-premises data sources across multiple gateway
connections. Accordingly, you must add all required data source definitions to the
same gateway.

Deploying a personal data gateway

If you have no access to an enterprise data gateway and you're the only person who
manages semantic models so you don't need to share data sources with others, you can
deploy a data gateway in personal mode. In the Gateway connection section, under You
have no personal gateways installed , select Install now. The personal data gateway has
several limitations as documented in On-premises data gateway (personal mode).

Unlike for an enterprise data gateway, you don't need to add data source definitions to
a personal gateway. Instead, you manage the data source configuration by using the
Data source credentials section in the semantic model settings, as the following
screenshot illustrates.

Accessing cloud data sources

Semantic models that use cloud data sources, such as Azure SQL DB, don't require a
data gateway if Power BI can establish a direct network connection to the source.
Accordingly, you can manage the configuration of these data sources by using the Data
source credentials section in the semantic model settings. As the following screenshot
shows, you don't need to configure a gateway connection.
 ７ Note

Each user can only have one set of credentials per data source, across all of the
semantic models they own, regardless of the workspaces where the semantic
models reside. And each semantic model can only have one owner. If your want to
update the credentials for a semantic model where you are not the semantic model
owner, you must first take over the semantic model by clicking on the Take Over
button on the semantic model settings page.

Accessing on-premises and cloud sources in the same

source query
A semantic model can get data from multiple sources, and these sources can reside on-
premises or in the cloud. However, a semantic model can only use a single gateway
connection, as mentioned earlier. While cloud data sources don't necessarily require a
gateway, a gateway is required if a semantic model connects to both on-premises and
cloud sources in a single mashup query. In this scenario, Power BI must use a gateway
for the cloud data sources as well. The following diagram illustrates how such a semantic
model accesses its data sources.
 ７ Note

If a semantic model uses separate mashup queries to connect to on-premises and

cloud sources, Power BI uses a gateway connection to reach the on-premises
sources and a direct network connection to the cloud sources. If a mashup query
merges or appends data from on-premises and cloud sources, Power BI switches to
the gateway connection even for the cloud sources.

Power BI semantic models rely on Power Query to access and retrieve source data. The
following mashup listing shows a basic example of a query that merges data from an
on-premises source and a cloud source.

Let

OnPremSource = Sql.Database("on-premises-db", "AdventureWorks"),

CloudSource = Sql.Databases("cloudsql.database.windows.net",
"AdventureWorks"),

TableData1 = OnPremSource{[Schema="Sales",Item="Customer"]}[Data],

TableData2 = CloudSource {[Schema="Sales",Item="Customer"]}[Data],

MergedData = Table.NestedJoin(TableData1, {"BusinessEntityID"},

TableData2, {"BusinessEntityID"}, "MergedData", JoinKind.Inner)
 in

MergedData

There are two options to configure a data gateway to support merging or appending
data from on-premises and cloud sources:

Add a data source definition for the cloud source to the data gateway in addition
to the on-premises data sources.
Enable the checkbox Allow user's cloud data sources to refresh through this
gateway cluster.

If you enable the checkbox Allow user's cloud data sources to refresh through this
gateway cluster in the gateway configuration, as in the screenshot above, Power BI can
use the configuration that the user defined for the cloud source under Data source
credentials in the semantic model settings. This can help to lower the gateway
configuration overhead. On the other hand, if you want to have greater control over the
connections that your gateway establishes, you shouldn't enable this checkbox. In this
case, you must add an explicit data source definition for every cloud source that you
want to support to your gateway. It's also possible to enable the checkbox and add
explicit data source definitions for your cloud sources to a gateway. In this case, the
gateway uses the data source definitions for all matching sources.

Configuring query parameters

The mashup or M queries you create by using Power Query can vary in complexity from
trivial steps to parameterized constructs. The following listing shows a small sample
mashup query that uses two parameters called SchemaName and TableName to access
a given table in an AdventureWorks database.

let

Source = Sql.Database("SqlServer01", "AdventureWorks"),

TableData = Source{[Schema=SchemaName,Item=TableName]}[Data]

in

TableData

７ Note

Query parameters are only supported for Import mode semantic models.
DirectQuery/LiveConnect mode does not support query parameter definitions.

To ensure that a parameterized semantic model accesses the correct data, you must
configure the mashup query parameters in the semantic model settings. You can also
update the parameters programmatically by using the Power BI REST API. The following
screenshot shows the user interface to configure the query parameters for a semantic
model that uses the above mashup query.
Refresh and dynamic data sources
A dynamic data source is a data source in which some or all of the information required
to connect can't be determined until Power Query runs its query, because the data is
generated in code or returned from another data source. Examples include: the instance
name and database of a SQL Server database; the path of a CSV file; or the URL of a web
service.

In most cases, Power BI semantic models that use dynamic data sources can't be
refreshed in the Power BI service. There are a few exceptions in which dynamic data
sources can be refreshed in the Power BI service, such as when using the RelativePath
and Query options with the Web.Contents M function. Queries that reference Power
Query parameters can also be refreshed.

To determine whether your dynamic data source can be refreshed, open the Data
Source Settings dialog in Power Query Editor, and then select Data Sources In Current
File. In the window that appears, look for the following warning message, as shown in
the following image:

７ Note

Some data sources may not be listed because of hand-authored queries.

If that warning is present in the Data Source Settings dialog that appears, then a
dynamic data source that can't be refreshed in the Power BI service is present.

Configure scheduled refresh

Establishing connectivity between Power BI and your data sources is by far the most
challenging task in configuring a data refresh. The remaining steps are relatively
straightforward and include setting the refresh schedule and enabling refresh failure
notifications. For step-by-step instructions, see the how-to guide Configuring scheduled
refresh.

Setting a refresh schedule

The Scheduled refresh section is where you define the frequency and time slots to
refresh a semantic model. As mentioned earlier, you can configure up to eight daily time
slots if your semantic model is on shared capacity, or 48 time slots on Power BI
Premium. The following screenshot shows a refresh schedule on a twelve-hour interval.
Having configured a refresh schedule, the semantic model settings page informs you
about the next refresh time, as in the screenshot above. If you want to refresh the data
sooner, such as to test your gateway and data source configuration, perform an on-
demand refresh by using the Refresh Now option in the semantic model menu in the
nav pane. On-demand refreshes don't affect the next scheduled refresh time.

 Tip

Power BI does not have a monthly refresh interval option. However, you can use
Power Automate to create a custom refresh interval that occurs monthly, as
described in the following Power BI blog post .

Note also that the configured refresh time might not be the exact time when Power BI
starts the next scheduled process. Power BI starts scheduled refreshes on a best effort
basis. The target is to initiate the refresh within 15 minutes of the scheduled time slot,
but a delay of up to one hour can occur if the service can't allocate the required
resources sooner.

７ Note

Power BI deactivates your refresh schedule after four consecutive failures or when
the service detects an unrecoverable error that requires a configuration update,
such as invalid or expired credentials. It is not possible to change the consecutive
failures threshold.

Getting refresh failure notifications

By default, Power BI sends refresh failure notifications to the semantic model owner
through email, so that they can act in a timely manner should refresh issues occur. If the
owner has the Power BI app on their mobile device, they will also get the failure
notification there. Power BI also sends an email notification when the service disables a
scheduled refresh due to consecutive failures. Microsoft recommends that you leave the
checkbox Send refresh failure notification emails semantic model owner enabled.

It's also a good idea to specify additional recipients for scheduled refresh failure
notifications by using the Email these contacts when the refresh fails textbox. Specified
recipients receive refresh failure notifications via email and push notifications to the
mobile app, just like the semantic model owner does. Specified recipients might include
a colleague taking care of your semantic models while you are on vacation, or the email
alias of your support team taking care of refresh issues for your department or
organization. Sending refresh failure notifications to others in addition to the semantic
model owner helps ensure that issues get noticed and addressed in a timely manner.

７ Note

Push notifications to the mobile apps do not support group aliases.

Note that Power BI not only sends notifications on refresh failures but also when the
service pauses a scheduled refresh due to inactivity. After two months, when no user has
visited any dashboard or report built on the semantic model, Power BI considers the
semantic model inactive. In this situation, Power BI sends an email message to the
semantic model owner indicating that the service paused the refresh schedule for the
semantic model. See the following screenshot for an example of such a notification.
To resume scheduled refresh, visit a report or dashboard built using this semantic model
or manually refresh the semantic model using the Refresh Now option.

７ Note

Sending refresh notifications to external users is not supported. The recipients you
specify in the Email these users when the refresh fails textbox must have accounts
in your Microsoft Entra tenant. This limitation applies to both semantic model
refresh and dataflow refresh.

Checking refresh status and history

In addition to failure notifications, it's a good idea to check your semantic models
periodically for refresh errors. A quick way is to view the list of semantic models in a
workspace. Semantic models with errors show a small warning icon. Select the warning
icon to obtain additional information, as in the following screenshot. For more
information about troubleshooting specific refresh errors, see Troubleshooting refresh
scenarios.

The warning icon helps to indicate current semantic model issues, but it's also a good
idea to check the refresh history occasionally. As the name implies, the refresh history
enables you to review the success or failure status of past synchronization cycles. For
example, a gateway administrator might have updated an expired set of database
credentials. As you can see in the following screenshot, the refresh history shows when
an affected refresh started working again.
 ７ Note

You can find a link to display the refresh history in the semantic model settings. You
can also retrieve the refresh history programmatically by using the Power BI REST
API. By using a custom solution, you can monitor the refresh history of multiple
semantic models in a centralized way.

Automatic page refresh

Automatic page refresh works at a report page level, and allows report authors to set a
refresh interval for visuals in a page that is only active when the page is being
consumed. Automatic page refresh is only available for DirectQuery data sources. The
minimum refresh interval depends on which type of workspace the report is published
in, and the capacity admin settings for Premium workspaces and embedded workspaces.

Learn more about automatic page refresh in the automatic page refresh article.

Semantic model refresh history

Refresh attempts for Power BI semantic models may not always go smoothly, or may
take longer than expected. You can use the Refresh history page to help you diagnose
why a refresh may not have happened as you expected.

Power BI automatically makes multiple attempts to refresh a semantic model if it

experiences a refresh failure. Without insight into refresh history activities, it may just
seem like a refresh is taking longer than expected. With the Refresh history page, you
can see those failed attempts and gain insight into the reason for the failure.

The following screenshot shows a failed refresh, with details about each time Power BI
automatically attempted to successfully complete the refresh.

You can also see when Power BI succeeds in when previous attempts failed, as shown in
the following image, which reveals that Power BI succeeded only after three previous
failures. Notice the successful data refresh and query cache share the same index
number, indicating they both were successful on the fourth attempt.
You can select the Show link beside a failure to get more information about the failed
refresh attempt, which can help with troubleshooting the issue.

In addition, each Power BI refresh attempt is divided into two operations:

Data – Load data into the semantic model

Query Cache – Premium query caches and/or dashboard tiles refresh

The following images show how Refresh history separates those operations, and
provides information about each.

Significant use of dashboard tiles or premium caching can increase refresh duration,
since either can queue many queries after each refresh. You can either reduce the
number of dashboards or disable automatic cache refresh setting to help reduce the
number of queries.

The data and query cache phases are independent of each other, but run in sequence.
The data refresh runs first and when that succeeds, the query cache refresh runs. If the
data refresh fails, the query refresh is not initiated. It's possible that the data refresh can
run successfully, but the query cache refresh fails.

Refreshes made using the XMLA Endpoint won't show attempt details in the Refresh
history window.

Refresh cancellation
Stopping a semantic model refresh is useful when you want to stop a refresh of a large
semantic model during peak time. Use the refresh cancellation feature to stop refreshing
semantic models that reside on Premium, Premium Per User (PPU) or Power BI
Embedded capacities.

To cancel a semantic model refresh, you need to be a contributor, member or an admin

of the semantic model's workspace. Semantic model refresh cancellation only works
with semantic models that use import mode or composite mode.

７ Note

Semantic models created as part of datamarts aren't supported.

To start a refresh go to the semantic model you want to refresh, and select Refresh now.
To stop a refresh follow these steps:

1. Go to the semantic model that's refreshing and select Cancel refresh.

2. In the Cancel refresh pop-up window, select Yes.

Best practices
Checking the refresh history of your semantic models regularly is one of the most
important best practices you can adopt to ensure that your reports and dashboards use
current data. If you discover issues, address them promptly and follow up with data
source owners and gateway administrators if necessary.

In addition, consider the following recommendations to establish and maintain reliable

data refresh processes for your semantic models:

Schedule your refreshes for less busy times, especially if your semantic models are
on Power BI Premium. If you distribute the refresh cycles for your semantic models
across a broader time window, you can help to avoid peaks that might otherwise
overtax available resources. Delays starting a refresh cycle are an indicator of
resource overload. If a Premium capacity is exhausted, Power BI might even skip a
refresh cycle.
Keep refresh limits in mind. If the source data changes frequently or the data
volume is substantial, consider using DirectQuery/LiveConnect mode instead of
Import mode if the increased load at the source and the impact on query
performance are acceptable. Avoid constantly refreshing an Import mode semantic
model. However, DirectQuery/LiveConnect mode has several limitations, such as a
one-million-row limit for returning data and a 225-seconds response time limit for
running queries, as documented in Use DirectQuery in Power BI Desktop. These
limitations might require you to use Import mode nonetheless. For large data
volumes, consider the use of aggregations in Power BI.
Verify that your semantic model refresh time doesn't exceed the maximum refresh
duration. Use Power BI Desktop to check the refresh duration. If it takes more than
2 hours, consider moving your semantic model to Power BI Premium. Your
semantic model might not be refreshable on shared capacity. Also consider using
Incremental refresh for semantic models that are larger than 1 GB or take several
hours to refresh.
Optimize your semantic models to include only those tables and columns that your
reports and dashboards use. Optimize your mashup queries and, if possible, avoid
dynamic data source definitions and expensive DAX calculations. Specifically avoid
DAX functions that test every row in a table because of the high memory
consumption and processing overhead.
Apply the same privacy settings as in Power BI Desktop to ensure that Power BI can
generate efficient source queries. Keep in mind that Power BI Desktop does not
publish privacy settings. You must manually reapply the settings in the data source
definitions after publishing your semantic model.
Limit the number of visuals on your dashboards, especially if you use row-level
security (RLS). As explained earlier in this article, an excessive number of dashboard
tiles can significantly increase the refresh duration.
Use a reliable enterprise data gateway deployment to connect your semantic
models to on-premises data sources. If you notice gateway-related refresh failures,
such as gateway unavailable or overloaded, follow up with gateway administrators
to either add additional gateways to an existing cluster or deploy a new cluster
(scale up versus scale out).
Use separate data gateways for Import semantic models and
DirectQuery/LiveConnect semantic models so that the data imports during
scheduled refresh don't impact the performance of reports and dashboards on top
of DirectQuery/LiveConnect semantic models, which query the data sources with
each user interaction.
Ensure that Power BI can send refresh failure notifications to your mailbox. Spam
filters might block the email messages or move them into a separate folder where
 you might not notice them immediately.

Related content
Configuring scheduled refresh
Tools for troubleshooting refresh issues
Troubleshooting refresh scenarios

More questions? Try asking the Power BI Community

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Managing query refresh in Power BI
Article • 01/23/2024

With Power BI, you can connect to many different types of data sources and shape the
data to meet your needs. The connections and transformations are stored in queries,
which are by default refreshed either by manual or automatic refresh of the report in the
Service.

Managing loading of queries

In many situations, it makes sense to break down your data transformations in multiple
queries. One popular example is merging where you merge two queries into one to
essentially do a join. In this situation, some queries aren't relevant to load into Power BI
Desktop because they're intermediate steps, while they're still required for your data
transformations to work correctly. For these queries, you can make sure they aren't
loaded in Power BI Desktop. Unselect Enable load in the context menu of the query in
Power Query Editor:

You can also make this change in the Properties screen:

Excluding queries from refresh
For queries for which the source data isn't updated often or at all, it makes sense to not
have the queries included in the refresh of the report. In this scenario, you can exclude
queries from being refreshed when the report is refreshed. Unselect Include in report
refresh in the context menu of the query in Power Query Editor:

You can also make this change in the Properties screen:

 ７ Note

Any queries excluded from refresh are also excluded in automatic refresh in the
Power BI service.

Related content
Shape and combine data
Configuring scheduled refresh
Tools for troubleshooting refresh issues
Troubleshooting refresh scenarios

More questions? Try asking the Power BI Community

Configure scheduled refresh
Article • 11/10/2023

７ Note

After two months of inactivity, scheduled refresh on your semantic model is

paused. For more information, see Scheduled refresh later in this article.

This article describes the options available for scheduled refresh for the On-premises
data gateway (personal mode) and the On-premises data gateway. You specify refresh
options in the following areas of the Power BI service: Gateway connection, Data source
credentials, and Schedule refresh. We'll look at each in turn. For more information
about data refresh, including limitations on refresh schedules, see Data refresh.

To get to the Schedule refresh screen:

1. In the navigation pane, under Semantic models, select a semantic model.

2. Select Refresh > Schedule refresh.

Gateway connection
You'll see different options here depending on whether you have a personal gateway or
enterprise gateway online and available.

If no gateway is available, you'll see Gateway connection disabled. You'll also see a
message indicating how to install the personal gateway.

If you have a personal gateway configured and it's online, it's available to select. It
shows offline if it's not available.
You can also select the enterprise gateway if one is available for you. You only see an
enterprise gateway available if your account is listed in the Users tab of the data source
configured for a given gateway.

Data source credentials

Power BI Gateway - Personal

If you're using the personal gateway to refresh data, you must supply the credentials to
connect to the back-end data source. If you connected to an app from an online service,
the credentials you entered to connect are carried over for scheduled refresh.

You're only required to sign in to a data source the first time you use refresh on that
semantic model. Once entered, those credentials are retained with the semantic model.

７ Note

For some authentication methods, if the password you use to sign into a data
source expires or is changed, you need to change it for the data source in Data
source credentials too.

If there's a problem, typically it's either the gateway is offline because it couldn't sign in
to Windows and start the service, or Power BI couldn't sign in to the data sources to
query for updated data. If refresh fails, check the semantic model settings. If the
gateway service is offline, Status is where you see the error. If Power BI can't sign into
the data sources, you see an error in Data Source Credentials.

On-premises data gateway

If you're using the on-premises data gateway to refresh data, you don't need to supply
credentials, as they're defined for the data source by the gateway administrator.

７ Note

When connecting to on-premises SharePoint for data refresh, Power BI supports

only Anonymous, Basic, and Windows (NTLM/Kerberos) authentication mechanisms.
Power BI does not support ADFS or any Forms-Based Authentication mechanisms
for data refresh of on-premises SharePoint data sources.

Scheduled refresh
The Scheduled refresh section is where you define the frequency and time slots to
refresh the semantic model. Some data sources don't require a gateway to be
configurable for refresh, while other data sources require a gateway.

In a Direct Query scenario, when a semantic model qualifies for performance

optimization, Refresh schedule will be moved to the Optimize performance section.

Set the Keep your data up to date slider to On to configure the settings.

７ Note

The target is to initiate the refresh within 15 minutes of the scheduled time slot, but
a delay of up to one hour can occur if the service can't allocate the required
resources sooner. Refresh can begin as early as five minutes before the scheduled
refresh time.
 ７ Note

After two months of inactivity, scheduled refresh on your semantic model is

paused. A semantic model is considered inactive when no user has visited any
dashboard or report built on the semantic model. When scheduled refresh is
paused, the semantic model owner is sent an email. The refresh schedule for the
semantic model is then displayed as disabled. To resume scheduled refresh, revisit
any dashboard or report built on the semantic model.

What's supported?

７ Note

Scheduled refresh will also get disabled automatically after four consecutive errors.

 Tip
 Power BI does not have a monthly refresh interval option. However, you can use
Power Automate to create a custom refresh interval that occurs monthly, as
described in the following Power BI blog post .

Certain semantic models are supported against different gateways for scheduled refresh.

Power BI Gateway - Personal

Power BI Desktop

All online data sources shown in Power BI Desktop's Get data and Power Query
Editor.
All on-premises data sources shown in Power BI Desktop's Get data and Power
Query Editor except for Hadoop file (HDFS) and Microsoft Exchange.

Excel

All online data sources shown in Power Query.

All on-premises data sources shown in Power Query except for Hadoop file (HDFS)
and Microsoft Exchange.
All online data sources shown in Power Pivot.
All on-premises data sources shown in Power Pivot except for Hadoop file (HDFS)
and Microsoft Exchange.

７ Note

In Excel 2016 and later, Launch Power Query Editor is available from Get Data in
the Data ribbon.

Power BI Gateway
For information about supported data sources, see Power BI data sources.

Troubleshooting
Sometimes refreshing data may not go as expected, typically due to an issue connected
with a gateway. See these gateway troubleshooting articles for tools and known issues.

Troubleshoot the On-premises data gateway

Troubleshoot the Power BI Gateway - Personal
Next steps
Data refresh in Power BI
Power BI Gateway - Personal
On-premises data gateway (personal mode)
Troubleshoot the On-premises data gateway
Troubleshoot the Power BI Gateway - Personal

More questions? Try asking the Power BI Community

Refresh summaries for Power BI
Article • 11/10/2023

The Power BI Refresh summary page, found in the Power BI Admin portal, provides
control and insight into your refresh schedules, capacities, and potential refresh
schedule overlaps for your Power BI Premium capacities. You can use the refresh
summary page to determine whether you should adjust refresh schedules, learn error
codes associated with refresh issues, and properly manage your data refresh scheduling.

The refresh summaries page has two views:

History. Displays the refresh summary history for Power BI Premium capacities for
which you're an administrator.
Schedule. Shows the schedule view for scheduled refresh, which also can uncover
issues with time slots that are oversubscribed.

You can also export information about a refresh event to a .csv file, which can provide
significant information and insight into refresh events or errors that can be impacting
the performance or completion of scheduled refresh events.

The following sections look at each of these views in turn.

Refresh history
You can select the History view by clicking on History in the refresh summaries page.

The History provides an overview of the outcomes of recently scheduled refreshes on

the capacities for which you have admin privilege. You can sort the view by any column
by selecting the column. You can choose to sort the view by the column selected by
ascending order, descending, or by using text filters.

In history view, the data associated with a given refresh is based on up 60 most recent
records for each scheduled refresh.

You can also export information for any scheduled refresh to a .csv file, which includes
detailed information including error messages for each refresh event. Exporting to a .csv
file lets you sort the file based on any of the columns, search for words, sort based on
error codes or owners, and so on. The following image shows an example exported .csv
file.

With the information in the exported file, you can review the capacity, duration, and any
error messages recorded for the instance of refresh.
Refresh schedule
You can select the Schedule view by selecting Schedule in refresh summaries. The
Schedule view displays scheduling information for the week, broken down into 30-
minute time slots.

The Schedule view is very useful in determining whether the refresh events scheduled
are properly spaced, allowing for all refreshes to complete without overlap, or whether
you have scheduled refresh events that are taking too long and creating resource
contention. If you find such resource contention, you should adjust your refresh
schedules to avoid the conflicts or overlap, so your scheduled refreshes can complete
successfully.

The Refresh time booked (minutes) column is a calculation of the average of up to 60

records for each associated semantic model. The numeric value for each 30-minute time
slot is the sum of minutes calculated for all scheduled refreshes scheduled to start on
the time slot and any scheduled refreshes set to start on the previous time slot, but
whose average duration overflows into the time slot that's selected.

The Refresh time available (minutes) column is a calculation of the minutes available for
refresh in each time slot, minus whatever refresh is already scheduled for that time slot.
For example, if your P2 subscription provides 80 concurrently running refreshes, you
have 80 30-minute slots, so 80 refreshes x 30 minutes each = 2,400 minutes available for
refresh in that time slot. If you have one refresh booked in that slot that takes 20
minutes, your Refresh time available (minutes) in that slot is 2,380 minutes (2,400 total
minutes available, minus 20 minutes already booked = 2,380 minutes still available).

You can select a time slot and then select the associated details button to see which
scheduled refresh events contribute to the refresh time booked, their owners, and how
long they take to complete.

Let's look at an example, to see how this works. The following dialog is displayed when
we select the 8:30 PM time slot for Sunday, and select details.

There are three scheduled refresh events occurring in this time slot.

Scheduled refresh #1 and #3 are both scheduled for this 8:30 PM time slot, which we
can determine by looking at the value in the Scheduled time slot column. Their average
durations are 4:39 and six seconds (0:06) respectively. All is good there.

However, scheduled refresh #2 is scheduled for the 8:00 PM time slot, but because it
takes an average of over 48 minutes to complete (seen in the Average duration
column), that refresh event overflows into the next 30-minute time slot.

That's not good. The administrator in this case should contact the owners of that
scheduled refresh instance and suggest they find a different time slot for that scheduled
refresh, or reschedule the other refreshes so there's no overlap, or find some other
solution to prevent such overlap.

Next steps
Data refresh in Power BI
Power BI Gateway - Personal
On-premises data gateway (personal mode)
Troubleshooting the On-premises data gateway
 Troubleshooting the Power BI Gateway - Personal

More questions? Try asking the Power BI Community

Enhanced refresh with the Power BI
REST API
Article • 01/08/2024

You can use any programming language that supports REST calls to do semantic model
refresh operations by using the Power BI Refresh Dataset REST API.

Optimized refresh for large and complex partitioned models is traditionally invoked with
programming methods that use TOM (Tabular Object Model), PowerShell cmdlets, or
TMSL (Tabular Model Scripting Language). However, these methods require long-
running HTTP connections that can be unreliable.

The Power BI Refresh Dataset REST API can carry out model refresh operations
asynchronously, so long-running HTTP connections from client applications aren't
necessary. Compared to standard refresh operations, enhanced refresh with the REST API
provides more customization options and the following features that are helpful for
large models:

Batched commits
Table and partition-level refresh
Applying incremental refresh policies
GET refresh details

Refresh cancellation

７ Note

Previously, enhanced refresh was called asynchronous refresh with REST API.
However, a standard refresh that uses the Refresh Dataset REST API also runs
asynchronously by its inherent nature.
Enhanced Power BI REST API refresh operations don't automatically refresh
tile caches. Tile caches refresh only when a user accesses a report.

Base URL
The base URL is in the following format:

HTTP
 https://api.powerbi.com/v1.0/myorg/groups/{groupId}/datasets/{datasetId}/ref
reshes

You can append resources and operations to the base URL based on parameters. In the
following diagram, Groups, Datasets, and Refreshes are collections. Group, Dataset, and
Refresh are objects.

Requirements
You need the following requirements to use the REST API:

A semantic model in Power BI Premium, Premium per user, or Power BI Embedded.

A group ID and dataset ID to use in the request URL.
Dataset.ReadWrite.All permission scope.

The number of refreshes is limited per the general limitations for API-based refreshes for
Pro and Premium models.

Authentication
All calls must authenticate with a valid Microsoft Entra ID OAuth 2 token in the
Authorization header. The token must meet the following requirements:

Be either a user token or an application service principal.

Have the audience correctly set to https://api.powerbi.com .
Be used by a user or application that has sufficient permissions on the model.

７ Note

REST API modifications don't change currently defined permissions for model
refreshes.

POST /refreshes
To do a refresh, use the POST verb on the /refreshes collection to add a new refresh
object to the collection. The Location header in the response includes the requestId .
Because the operation is asynchronous, a client application can disconnect and use the
requestId to check the status later if necessary.

The following code shows a sample request:

HTTP

POST https://api.powerbi.com/v1.0/myorg/groups/f089354e-8366-4e18-aea3-
4cb4a3a50b48/datasets/cfafbeb1-8037-4d0c-896e-a46fb27ff229/refreshes

The request body might resemble the following example:

JSON

{
"type": "Full",
"commitMode": "transactional",
"maxParallelism": 2,
"retryCount": 2,
"objects": [
{
"table": "DimCustomer",
"partition": "DimCustomer"
},
{
"table": "DimDate"
}
]
}

７ Note

The service accepts only one refresh operation at a time for a model. If there's a
current running refresh and another request is submitted, a 400 Bad Request HTTP
status code returns.

Parameters
To do an enhanced refresh operation, you must specify one or more parameters in the
request body. Specified parameters can specify the default or an optional value. When
the request specifies parameters, all other parameters apply to the operation with their
default values. If the request specifies no parameters, all parameters use their default
values, and a standard refresh operation occurs.
 ﾉ Expand table

Name Type Default Description

type Enum automatic The type of processing to perform. Types align

with the TMSL refresh command types: full ,
clearValues , calculate , dataOnly , automatic ,
and defragment . The add type isn't supported.

commitMode Enum transactional Determines whether to commit objects in

batches or only when complete. Modes are
transactional and partialBatch . When using
partialBatch the refresh operation doesn’t
occur within one transaction. Each command is
committed individually. If there’s a failure, the
model might be empty or include only a subset
of the data. To safeguard against failure and
keep the data that was in the model before the
operation started, execute the operation with
commitMode = transactional .

maxParallelism Int 10 Determines the maximum number of threads

that can run the processing commands in
parallel. This value aligns with the
MaxParallelism property that can be set in the
TMSL Sequence command or by using other
methods.

retryCount Int 0 Number of times the operation retries before

failing.

objects Array Entire model An array of objects to process. Each object

includes table when processing an entire table,
or table and partition when processing a
partition. If no objects are specified, the entire
model refreshes.

applyRefreshPolicy Boolean true If an incremental refresh policy is defined,

determines whether to apply the policy. Modes
are true or false . If the policy isn't applied,
the full process leaves partition definitions
unchanged, and fully refreshes all partitions in
the table.

If commitMode is transactional ,
applyRefreshPolicy can be true or false . If
commitMode is partialBatch ,
applyRefreshPolicy of true isn't supported,
and applyRefreshPolicy must be set to false .
 Name Type Default Description

effectiveDate Date Current date If an incremental refresh policy is applied, the

effectiveDate parameter overrides the current
date. If not specified, UTC is used to determine
the current day.

Response
JSON

202 Accepted

The response also includes a Location response-header field to point the caller to the
refresh operation that was created and accepted. The Location is the location of the
new resource the request created, which includes the requestId that some enhanced
refresh operations require. For example, in the following response, requestId is the last
identifier in the response 87f31ef7-1e3a-4006-9b0b-191693e79e9e .

JSON

x-ms-request-id: 87f31ef7-1e3a-4006-9b0b-191693e79e9e
Location: https://api.powerbi.com/v1.0/myorg/groups/f089354e-8366-4e18-aea3-
4cb4a3a50b48/datasets/cfafbeb1-8037-4d0c-896e-
a46fb27ff229/refreshes/87f31ef7-1e3a-4006-9b0b-191693e79e9e

GET /refreshes
Use the GET verb on the /refreshes collection to list historical, current, and pending
refresh operations.

The response body might look like the following example:

JSON

[
{
"requestId": "1344a272-7893-4afa-a4b3-3fb87222fdac",
"refreshType": "ViaEnhancedApi",
"startTime": "2020-12-07T02:06:57.1838734Z",
"endTime": "2020-12-07T02:07:00.4929675Z",
"status": "Completed",
"extendedStatus": "Completed"
},
 {
"requestId": "474fc5a0-3d69-4c5d-adb4-8a846fa5580b",
"startTime": "2020-12-07T01:05:54.157324Z",
"refreshType": "ViaEnhancedApi",
"endTime": "2020-12-07T01:05:57.353371Z",
"status": "Unknown"
}
{
"requestId": "85a82498-2209-428c-b273-f87b3a1eb905",
"refreshType": "ViaEnhancedApi",
"startTime": "2020-12-07T01:05:54.157324Z",
"endTime": "2020-12-07T01:05:57.353371Z",
"status": "Unknown",
"extendedStatus": "NotStarted"
}
]

７ Note

Power BI might drop requests if there are too many requests in a short period of
time. Power BI does a refresh, queues the next request, and drops all others. By
design, you can't query status on dropped requests.

Response properties

ﾉ Expand table

Name Type Description

requestId Guid The identifier of the refresh request. You need requestId to query for
individual refresh operation status or cancel an in-progress refresh
operation.

refreshType String OnDemand indicates the refresh was triggered interactively through the
Power BI portal.
Scheduled indicates that a model refresh schedule triggered the
refresh.
ViaApi indicates that an API call triggered the refresh.
ViaEnhancedApi indicates that an API call triggered an enhanced
refresh.

startTime String Date and time of refresh start.

endTime String Date and time of refresh end.

 Name Type Description

status String Completed indicates the refresh operation completed successfully.

Failed indicates the refresh operation failed.
Unknown indicates that the completion state can't be determined. With
this status, endTime is empty.
Disabled indicates that the refresh was disabled by selective refresh.
Cancelled indicates the refresh was canceled successfully.

extendedStatus String Augments the status property to provide more information.

７ Note

In Azure Analysis Services, the completed status result is succeeded . If you migrate
an Azure Analysis Services solution to Power BI, you might have to modify your
solutions.

Limit the number of refresh operations returned

The Power BI REST API supports limiting the requested number of entries in the refresh
history by using the optional $top parameter. If not specified, the default is all available
entries.

HTTP

GET
https://api.powerbi.com/v1.0/myorg/groups/{groupId}/datasets/{datasetId}/ref
reshes?$top={$top}

GET /refreshes/<requestId>
To check the status of a refresh operation, use the GET verb on the refresh object by
specifying the requestId . If the operation is in progress, status returns InProgress , as
in the following example response body:

JSON

{
"startTime": "2020-12-07T02:06:57.1838734Z",
"endTime": "2020-12-07T02:07:00.4929675Z",
"type": "Full",
"status": "InProgress",
"currentRefreshType": "Full",
 "objects": [
{
"table": "DimCustomer",
"partition": "DimCustomer",
"status": "InProgress"
},
{
"table": "DimDate",
"partition": "DimDate",
"status": "InProgress"
}
]
}

DELETE /refreshes/<requestId>
To cancel an in-progress enhanced refresh operation, use the DELETE verb on the
refresh object by specifying the requestId .

For example,

HTTP

DELETE https://api.powerbi.com/v1.0/myorg/groups/f089354e-8366-4e18-aea3-
4cb4a3a50b48/datasets/cfafbeb1-8037-4d0c-896e-
a46fb27ff229/refreshes/1344a272-7893-4afa-a4b3-3fb87222fdac

Considerations and limitations

The refresh operation has the following considerations and limitations:

Standard refresh operations

You can't cancel scheduled or on-demand manual model refreshes by using DELETE
/refreshes/<requestId> .

Scheduled and on-demand manual model refreshes don't support getting refresh
operation details by using GET /refreshes/<requestId> .
Get details and Cancel are new operations for enhanced refresh only. Standard
refresh doesn't support these operations.

Power BI Embedded
If capacity is paused manually in the Power BI portal or by using PowerShell, or a system
outage occurs, the status of any ongoing enhanced refresh operation remains
InProgress for a maximum of six hours. If the capacity resumes within six hours, the

refresh operation resumes automatically. If the capacity resumes after longer than six
hours, the refresh operation might return a timeout error. You must then restart the
refresh operation.

Semantic model eviction

Power BI uses dynamic memory management to optimize capacity memory. If the
model is evicted from memory during a refresh operation, the following error might
return:

JSON

{
"messages": [
{
"code": "0xC11C0020",
"message": "Session cancelled because it is connected to a
database that has been evicted to free up memory for other operations",
"type": "Error"
}
]
}

The solution is to rerun the refresh operation. To learn more about dynamic memory
management and model eviction, see Model eviction.

Refresh operation time limits

The maximum amount of time for a single refresh operation is five hours. If the refresh
operation doesn't successfully complete within the five-hour limit, and retryCount isn't
specified or is specified as 0 (the default) in the request body, a timeout error returns.

If retryCount specifies 1 or another number, a new refresh operation with a five-hour

limit starts. If this retry operation fails, the service continues to retry the refresh
operation up to the greatest number of retries that retryCount specifies, or the
enhanced refresh processing time limit of 24 hours from the beginning of the first
refresh request.

When you plan your enhanced model refresh solution with the Refresh Dataset REST
API, it's important to consider these time limits and the retryCount parameter. A
successful refresh completion can exceed five hours if an initial refresh operation fails
and retryCount specifies 1 or more.

For example, if you request a refresh operation with "retryCount": 1 , and the initial
retry operation fails four hours from the start time, a second refresh operation for that
request begins. If that second refresh operation succeeds in three hours, the total time
for successful execution of the refresh request is seven hours.

If refresh operations regularly fail, exceed the five-hour time limit, or exceed your
desired successful refresh operation time, consider reducing the amount of data being
refreshed from the data source. You can split refresh into multiple requests, for example
a request for each table. You can also specify partialBatch in the commitMode
parameter.

Code sample
For a C# code sample to get you started, see RestApiSample on GitHub.

To use the code sample:

1. Clone or download the repo.

2. Open the RestApiSample solution.
3. Find the line client.BaseAddress = … and provide your base URL.

The code sample uses service principal authentication.

See also
Power BI Refresh Dataset REST API
Use the Power BI REST APIs
Incremental refresh and real-time data
for semantic models
Article • 11/10/2023

Incremental refresh extends scheduled refresh operations by providing automated

partition creation and management for semantic model tables that frequently load new
and updated data. For most models, one or more tables contain transaction data that
changes often and can grow exponentially, like a fact table in a relational or star
database schema. An incremental refresh policy to partition the table, refreshing only
the most recent import partitions, and optionally using another DirectQuery partition for
real-time data can significantly reduce the amount of data that has to be refreshed. At
the same time, this policy ensures that the latest changes at the data source are
included in the query results.

With incremental refresh and real-time data:

Fewer refresh cycles for fast-changing data are needed. DirectQuery mode gets
the latest data updates as queries are processed, without requiring a high refresh
cadence.
Refreshes are faster. Only the most recent data that has changed needs to be
refreshed.
Refreshes are more reliable. Long-running connections to volatile data sources
aren't necessary. Queries to source data run faster, reducing potential for network
problems to interfere.
Resource consumption is reduced. Less data to refresh reduces overall
consumption of memory and other resources in both Power BI and data source
systems.
Large semantic models are enabled. Semantic models with potentially billions of
rows can grow without the need to fully refresh the entire model with each refresh
operation.
Setup is easy. Incremental refresh policies are defined in Power BI Desktop with
just a few tasks. When Power BI Desktop publishes the report, the service
automatically applies those policies with each refresh.

When you publish a Power BI Desktop model to the service, each table in the new
model has a single partition. That single partition contains all rows for that table. If the
table is large, say with tens of millions of rows or more, a refresh for that table can take
a long time and consume an excessive amount of resources.
With incremental refresh, the service dynamically partitions and separates data that
needs to be refreshed frequently from data that can be refreshed less frequently. Table
data is filtered by using Power Query date/time parameters with the reserved, case-
sensitive names RangeStart and RangeEnd . When you configure incremental refresh in
Power BI Desktop, these parameters are used to filter only a small period of data that's
loaded into the model. When Power BI Desktop publishes the report to the Power BI
service, with the first refresh operation the service creates incremental refresh and
historical partitions, and optionally a real-time DirectQuery partition based on the
incremental refresh policy settings. The service then overrides the parameter values to
filter and query data for each partition based on date/time values for each row.

With each subsequent refresh, the query filters return only those rows within the refresh
period dynamically defined by the parameters. Those rows with a date/time within the
refresh period are refreshed. Rows with a date/time no longer within the refresh period
then become part of the historical period, which isn't refreshed. If a real-time
DirectQuery partition is included in the incremental refresh policy, its filter is also
updated so that it picks up any changes that occur after the refresh period. Both the
refresh and historical periods are rolled forward. As new incremental refresh partitions
are created, refresh partitions no longer in the refresh period become historical
partitions. Over time, historical partitions become less granular as they're merged
together. When a historical partition is no longer in the historical period defined by the
policy, it's removed from the model entirely. This behavior is known as a rolling window
pattern.

The beauty of incremental refresh is that the service handles all of it for you based on
the incremental refresh policies you define. In fact, the process and partitions created
from it aren't visible in the service. In most cases, a well-defined incremental refresh
policy is all that's necessary to significantly improve model refresh performance.
However, the real-time DirectQuery partition is only supported for models in Premium
capacities. Power BI Premium also enables more advanced partition and refresh
scenarios through the XML for Analysis (XMLA) endpoint.

Requirements
The next sections describe the supported plans and data sources.

Supported plans
Incremental refresh is supported for Power BI Premium, Premium per user, Power BI Pro,
and Power BI Embedded models.

Getting the latest data in real time with DirectQuery is only supported for Power BI
Premium, Premium per user, and Power BI Embedded models.

Supported data sources

Incremental refresh and real-time data works best for structured, relational data sources
like SQL Database and Azure Synapse, but can also work for other data sources. In any
case, your data source must support the following:

Date filtering - The data source must support some mechanism to filter data by date.
For a relational source this is typically a date column of date/time or integer data type
on the target table. The RangeStart and RangeEnd parameters, which must be date/time
data type, filter table data based on the date column. For date columns of integer
surrogate keys in the form of yyyymmdd , you can create a function that converts the
date/time value in the RangeStart and RangeEnd parameters to match the integer
surrogate keys of the date column. To learn more, see Configure incremental refresh -
Convert DateTime to integer.

For other data sources, the RangeStart and RangeEnd parameters must be passed to the
data source in some way that enables filtering. For file-based data sources where files
and folders are organized by date, the RangeStart and RangeEnd parameters can be
used to filter the files and folders to select which files to load. For web-based data
sources the RangeStart and RangeEnd parameters can be integrated into the HTTP
request. For example, the following query can be used for incremental refresh of the
traces from an AppInsights instance:

Power Query M

let
strRangeStart = DateTime.ToText(RangeStart,[Format="yyyy-MM-
dd'T'HH:mm:ss'Z'", Culture="en-US"]),
strRangeEnd = DateTime.ToText(RangeEnd,[Format="yyyy-MM-
dd'T'HH:mm:ss'Z'", Culture="en-US"]),
Source =
Json.Document(Web.Contents("https://api.applicationinsights.io/v1/apps/<app-
guid>/query",
[Query=[#"query"="traces
 | where timestamp >= datetime(" & strRangeStart &")
| where timestamp < datetime("& strRangeEnd &")
",#"x-ms-app"="AAPBI",#"prefer"="ai.response-
thinning=true"],Timeout=#duration(0,0,4,0)])),
TypeMap = #table(
{ "AnalyticsTypes", "Type" },
{
{ "string", Text.Type },
{ "int", Int32.Type },
{ "long", Int64.Type },
{ "real", Double.Type },
{ "timespan", Duration.Type },
{ "datetime", DateTimeZone.Type },
{ "bool", Logical.Type },
{ "guid", Text.Type },
{ "dynamic", Text.Type }
}),
DataTable = Source[tables]{0},
Columns = Table.FromRecords(DataTable[columns]),
ColumnsWithType = Table.Join(Columns, {"type"}, TypeMap ,
{"AnalyticsTypes"}),
Rows = Table.FromRows(DataTable[rows], Columns[name]),
Table = Table.TransformColumnTypes(Rows, Table.ToList(ColumnsWithType,
(c) => { c{0}, c{3}}))
in
Table

When incremental refresh is configured, a Power Query expression that includes a

date/time filter based on the RangeStart and RangeEnd parameters is executed against
the data source. If the filter is specified in a query step after the initial source query, it's
important that query folding combines the initial query step with the steps reference the
RangeStart and RangeEnd parameters. For example, in the following query expression,
the Table.SelectRows will fold because it immediately follows the Sql.Database step,
and SQL Server supports folding:

Power Query M

let
Source = Sql.Database("dwdev02","AdventureWorksDW2017"),
Data = Source{[Schema="dbo",Item="FactInternetSales"]}[Data],
#"Filtered Rows" = Table.SelectRows(Data, each [OrderDateKey] >=
Int32.From(DateTime.ToText(RangeStart,[Format="yyyyMMdd"]))),
#"Filtered Rows1" = Table.SelectRows(#"Filtered Rows", each [OrderDateKey]
< Int32.From(DateTime.ToText(RangeEnd,[Format="yyyyMMdd"])))

in
#"Filtered Rows1"

There's no requirement the final query support folding. For example in the following
expression, we use a non-folding NativeQuery but integrate the RangeStart and
RangeEnd parameters directly into SQL:

Power Query M

let
Query = "select * from dbo.FactInternetSales where OrderDateKey >= '"&
Text.From(Int32.From( DateTime.ToText(RangeStart,"yyyyMMdd") )) &"' and
OrderDateKey < '"& Text.From(Int32.From(
DateTime.ToText(RangeEnd,"yyyyMMdd") )) &"' ",
Source = Sql.Database("dwdev02","AdventureWorksDW2017"),
Data = Value.NativeQuery(Source, Query, null, [EnableFolding=false])
in
Data

However, if the incremental refresh policy includes getting real-time data with
DirectQuery, non-folding transformations can't be used. If it’s a pure import mode
policy without real-time data, the query mashup engine might compensate and apply
the filter locally, which requires retrieving all rows for the table from the data source.
This can cause incremental refresh to be slow, and the process can run out of resources
either in the Power BI service or in an On-premises Data Gateway - effectively defeating
the purpose of incremental refresh.

Because support for query folding is different for different types of data sources,
verification should be performed to ensure the filter logic is included in the queries
being run against the data source. In most cases, Power BI Desktop attempts to perform
this verification for you when defining the incremental refresh policy. For SQL-based
data sources such as SQL Database, Azure Synapse, Oracle, and Teradata, this
verification is reliable. However, other data sources may be unable to verify without
tracing the queries. If Power BI Desktop is unable to confirm the queries, a warning is
shown in the Incremental refresh policy configuration dialog.

If you see this warning and want to verify the necessary query folding is occurring, use
the Power Query Diagnostics feature, or trace queries by using a tool supported by the
data source, such as SQL Profiler. If query folding isn't occurring, verify the filter logic is
included in the query being passed to the data source. If not, it's likely the query
includes a transformation that prevents folding.
Before configuring your incremental refresh solution, be sure to thoroughly read and
understand Query folding guidance in Power BI Desktop and Power Query query
folding. These articles can help you determine whether your data source and queries
support query folding.

Single data source

When you configure incremental refresh and real-time data by using Power BI Desktop,
or configure an advanced solution by using Tabular Model Scripting Language (TMSL) or
Tabular Object Model (TOM) through the XMLA endpoint, all partitions, whether import
or DirectQuery, must query data from a single source.

Other data source types

By using more custom query functions and query logic, incremental refresh can be used
with other types of data sources if filters based on RangeStart and RangeEnd can be
passed in a single query, like with data sources such as Excel workbook files stored in a
folder, files in SharePoint, and RSS feeds. Keep in mind these are advanced scenarios
that require further customization and testing beyond what is described here. Be sure to
check the Community section later in this article for suggestions on how you can find
more information about using incremental refresh for unique scenarios.

Time limits
Regardless of incremental refresh, Power BI Pro models have a refresh time limit of two
hours and don't support getting real-time data with DirectQuery. For models in a
Premium capacity, the time limit is five hours. Refresh operations are process and
memory intensive. A full refresh operation can use as much as double the amount of
memory required by the model alone, because the service maintains a snapshot of the
model in memory until the refresh operation is complete. Refresh operations can also be
process intensive, consuming a significant amount of available CPU resources. Refresh
operations must also rely on volatile connections to data sources, and the ability of
those data source systems to quickly return query output. The time limit is a safeguard
to limit over-consumption of your available resources.

７ Note
 With Premium capacities, refresh operations performed through the XMLA
endpoint have no time limit. To learn more, see Advanced incremental refresh with
the XMLA endpoint.

Because incremental refresh optimizes refresh operations at the partition level in the
model, resource consumption can be significantly reduced. At the same time, even with
incremental refresh, unless they go through the XMLA endpoint, refresh operations are
bound by those same two-hour and five-hour limits. An effective incremental refresh
policy not only reduces the amount of data processed with a refresh operation, but also
reduces the amount of unnecessary historical data stored in your model.

Queries can also be limited by a default time limit for the data source. Most relational
data sources allow overriding time limits in the Power Query M expression. For example,
the expression below uses the SQL Server data-access function to set CommandTimeout
to 2 hours. Each period defined by the policy ranges submits a query observing the
command timeout setting:

Power Query M

let
Source = Sql.Database("myserver.database.windows.net", "AdventureWorks",
[CommandTimeout=#duration(0, 2, 0, 0)]),
dbo_Fact = Source{[Schema="dbo",Item="FactInternetSales"]}[Data],
#"Filtered Rows" = Table.SelectRows(dbo_Fact, each [OrderDate] >=
RangeStart and [OrderDate] < RangeEnd)
in
#"Filtered Rows"

For very large models in Premium capacities that likely contain billions of rows, the initial
refresh operation can be bootstrapped. Bootstrapping allows the service to create table
and partition objects for the model, but doesn't load and process data into any of the
partitions. By using SQL Server Management Studio, you can set partitions to be
processed individually, sequentially, or in parallel, to both reduce the amount of data
returned in a single query, and also bypass the five-hour time limit. To learn more, see
Advanced incremental refresh - Prevent timeouts on initial full refresh.

Current date and time

The current date and time is based on the system date at the time of refresh. If
scheduled refresh is enabled for the model in the service, the specified time zone is
taken into account when determining the current date and time. Both individual and
scheduled refreshes through the service observe the time zone if available. For example,
a refresh that occurs at 8:00 PM Pacific Time (US and Canada) with a time zone specified
determines the current date and time based on Pacific Time, not Coordinated Universal
Time (UTC), which would return the next day. Refresh operations not invoked through
the Power BI service, such as the TMSL refresh command, don't consider the scheduled
refresh time zone.

Configure incremental refresh and real-time

data
This section describes important concepts of configuring incremental refresh and real-
time data. When you're ready for more detailed step-by-step instructions, see Configure
incremental refresh and real-time data for models.

Configuring incremental refresh is done in Power BI Desktop. For most models, only a
few tasks are required. However, keep the following points in mind:

After publishing to the Power BI service, you can't publish the same model again
from Power BI Desktop. Republishing removes any existing partitions and data
already in the model. If you're publishing to a Premium capacity, subsequent
 metadata schema changes can be made with tools such as the open-source ALM
Toolkit, or by using TMSL. To learn more, see Advanced incremental refresh -
Metadata-only deployment.
After publishing to the Power BI service, you can't download the model back as a
.pbix to Power BI Desktop. Because models in the service can grow so large, it's
impractical to download and open them on a typical desktop computer.
When getting real-time data with DirectQuery, you can't publish the model to a
non-Premium workspace. Incremental refresh with real-time data is only supported
with Power BI Premium.

Create parameters
To configure incremental refresh in Power BI Desktop, you first create two Power Query
date/time parameters with the reserved, case-sensitive names RangeStart and RangeEnd .
These parameters, defined in the Manage Parameters dialog in Power Query Editor, are
initially used to filter the data loaded into the Power BI Desktop model table to include
only those rows with a date/time within that period. RangeStart represents the oldest,
or earliest date/time, and RangeEnd represents the newest, or latest date/time. After the
model is published to the service, RangeStart and RangeEnd are overridden
automatically by the service to query data defined by the refresh period specified in the
incremental refresh policy settings.

For example, the FactInternetSales data source table averages 10,000 new rows per day.
To limit the number of rows initially loaded into the model in Power BI Desktop, specify
a two-day period between RangeStart and RangeEnd .
Filter data
With the RangeStart and RangeEnd parameters defined, you apply custom date filters on
your table's date column. The filters you apply select a subset of data that's loaded into
the model when you select Apply.
With our FactInternetSales example, after creating filters based on the parameters and
applying steps, two days of data (roughly 20,000 rows) are loaded into the model.

Define policy
After filters have been applied and a subset of data has been loaded into the model, you
define an incremental refresh policy for the table. After the model is published to the
service, the policy is used by the service to create and manage table partitions and
perform refresh operations. To define the policy, you use the Incremental refresh and
real-time data dialog box to specify both required and optional settings.
Table

The Select table listbox defaults to the table you selected in Data view. Enable
incremental refresh for the table with the slider. If the Power Query expression for the
table doesn't include a filter based on the RangeStart and RangeEnd parameters, the
toggle isn't available.

Required settings
The Archive data starting before refresh date setting determines the historical period in
which rows with a date/time in that period are included in the model, plus rows for the
current incomplete historical period, plus rows in the refresh period up to the current
date and time.

For example, if you specify five years, the table stores the last five whole years of
historical data in year partitions. The table will also include rows for the current year in
quarter, month, or day partitions, up to and including the refresh period.

For models in Premium capacities, backdated historical partitions can be selectively

refreshed at a granularity determined by this setting. To learn more, see Advanced
incremental refresh - Partitions.

The Incrementally refresh data starting before refresh date setting determines the
incremental refresh period in which all rows with a date/time in that period are included
in the refresh partitions and refreshed with each refresh operation.

For example, if you specify a refresh period of three days, with each refresh operation,
the service overrides the RangeStart and RangeEnd parameters to create a query for
rows with a date/time within a three-day period, with the beginning and ending
dependent on the current date and time. Rows with a date/time in the last three days up
to the current refresh operation time are refreshed. With this type of policy, you can
expect our FactInternetSales model table in the service, which averages 10,000 new rows
per day, to refresh roughly 30,000 rows with each refresh operation.

Specify a period that includes only the minimum number of rows required to ensure
accurate reporting. When you define policies for more than one table, the same
RangeStart and RangeEnd parameters must be used even if different store and refresh
periods are defined for each table.

Optional settings

The Get the latest data in real time with DirectQuery (Premium only) setting enables
fetching the latest changes from the selected table at the data source beyond the
incremental refresh period by using DirectQuery. All rows with a date/time later than the
incremental refresh period are included in a DirectQuery partition and fetched from the
data source with every model query.

For example, if this setting is enabled, with each refresh operation, the service still
overrides the RangeStart and RangeEnd parameters to create a query for rows with a
date/time after the refresh period, with the beginning dependent on the current date
and time. Rows with a date/time after the current refresh operation time are also
included. With this type of policy, the FactInternetSales model table in the service
includes the latest data updates.
The Only refresh complete days setting ensures all rows for the entire day are included
in the refresh operation. This setting is optional unless you enable the Get the latest
data in real time with DirectQuery (Premium only) setting. For example, say your
refresh is scheduled to run at 4:00 AM every morning. If new rows of data appear in the
data source table during those four hours between midnight and 4:00 AM, you don't
want to account for them. Some business metrics like barrels per day in the oil and gas
industry make no sense with partial days. Another example is refreshing data from a
financial system where data for the previous month is approved on the twelfth calendar
day of the month. You could set the refresh period to one month and schedule the
refresh to run on the twelfth day of the month. With this option selected, it would, for
example, refresh January data on February 12.

Keep in mind, unless scheduled refresh is configured for a non-UTC time zone, refresh
operations in the service run under UTC time, which can determine the effective date
and complete periods.

The Detect data changes setting enables even more selective refresh. You can select a
date/time column used to identify and refresh only those days where the data has
changed. This setting assumes such a column exists in the data source, which is typically
for auditing purposes. This column shouldn't be the same column used to partition the
data with the RangeStart and RangeEnd parameters. The maximum value of this column
is evaluated for each of the periods in the incremental range. If it hasn't changed since
the last refresh, there's no need to refresh the period, which could potentially further
reduce the days incrementally refreshed from three to one.

The current design requires that the column to detect data changes is persisted and
cached into memory. The following techniques can be used to reduce cardinality and
memory consumption:

Persist only the maximum value of the column at time of refresh, perhaps by using
a Power Query function.
Reduce the precision to an acceptable level, given your refresh-frequency
requirements.
Define a custom query for detecting data changes by using the XMLA endpoint,
and avoid persisting the column value altogether.

In some cases, enabling the Detect data changes* option can be further enhanced. For
example, you may want to avoid persisting a last-update column in the in-memory
cache, or enable scenarios where a configuration/instruction table is prepared by
extract-transform-load (ETL) processes for flagging only those partitions that need to be
refreshed. In cases like these, for Premium capacities, use TMSL and/or the TOM to
override the detect data changes behavior. To learn more, see Advanced incremental
refresh - Custom queries for detect data changes.

Publish
After configuring the incremental refresh policy, you publish the model to the service.
When publishing is complete, you can perform the initial refresh operation on the
model.

７ Note

Semantic models with an incremental refresh policy to get the latest data in real
time with DirectQuery can only be published to a Premium workspace.

For models published to workspaces assigned to Premium capacities, if you think the
model will grow beyond 1 GB, you can improve refresh operation performance and
ensure the model doesn't max out size limits by enabling Large model storage format
before performing the first refresh operation in the service. To learn more, see Large
models in Power BI Premium.

） Important

After Power BI Desktop publishes the model to the service, you can't download that
.pbix back.

Refresh
After publishing to the service, you perform an initial refresh operation on the model.
This refresh should be an individual (manual) refresh so you can monitor progress. The
initial refresh operation can take quite a while to complete. Partitions must be created,
historical data loaded, objects such as relationships and hierarchies built or rebuilt, and
calculated objects recalculated.

Subsequent refresh operations, either individual or scheduled, are much faster because
only the incremental refresh partitions are refreshed. Other processing operations must
still occur, like merging partitions and recalculation, but it usually takes much less time
than the initial refresh.

Automatic report refresh

For reports that use a model with an incremental refresh policy to get the latest data in
real time with DirectQuery, it's a good idea to enable automatic page refresh at a fixed
interval or based on change detection so that the reports include the latest data without
delay. To learn more, see Automatic page refresh in Power BI.

Advanced incremental refresh

If your model is on a Premium capacity with an XMLA endpoint enabled, incremental
refresh can be further extended for advanced scenarios. For example, you can use SQL
Server Management Studio to view and manage partitions, bootstrap the initial refresh
operation, or refresh backdated historical partitions. To learn more, see Advanced
incremental refresh with the XMLA endpoint.

Community
Power BI has a vibrant community where MVPs, BI pros, and peers share expertise in
discussion groups, videos, blogs, and more. When learning about incremental refresh,
see these resources:

Power BI Community
Search "Power BI incremental refresh" on Bing
Search "Incremental refresh for files" on Bing
Search "Keep existing data using incremental refresh" on Bing

Next steps
Configure incremental refresh for semantic models
Advanced incremental refresh with the XMLA endpoint
Troubleshoot incremental refresh
Incremental refresh for dataflows
Configure incremental refresh and real-
time data
Article • 04/25/2024

This article describes how to configure incremental refresh and real-time data for
semantic models. To learn about configuring incremental refresh for dataflows, see
Premium features of dataflows - Incremental refresh.

Configuring incremental refresh includes creating RangeStart and RangeEnd parameters,

applying filters, and defining an incremental refresh policy. After publishing to the Power
BI service, you'll perform an initial refresh operation on the model. The initial refresh
operation, and subsequent refresh operations apply the incremental refresh policy you
defined. Before completing these steps, be sure you fully understand the functionality
described in Incremental refresh and real-time data for semantic models.

Create parameters
In this task, you'll use Power Query Editor to create RangeStart and RangeEnd
parameters with default values. The default values apply only when filtering the data to
be loaded into the model in Power BI Desktop. The values you enter should include only
a small amount of the most recent data from your data source. When published to the
service, these time range values are overridden by the incremental refresh policy. That is,
the policy creates windows of incoming data, one after another.

1. In Power BI Desktop, select Transform data on the Home ribbon to open Power
Query Editor.

2. Select the Manage Parameters dropdown and then choose New Parameter.

3. In the Name field, enter RangeStart (case-sensitive). In the Type field, select
Date/Time from the dropdown. In the Current Value field, enter a start date and
time value.
4. Select New to create a second parameter named RangeEnd. In the Type field,
select Date/Time, and then in the Current Value field enter an end date and time
value. Select OK.
Now that you've defined the RangeStart and RangeEnd parameters, you'll filter the data
to be loaded into the model based on those parameters.

Filter data

７ Note

Before continuing with this task, verify your source table has a date column of
Date/Time data type. If it doesn’t have a Date/Time column, but it has a date
column of integer surrogate keys in the form of yyyymmdd , follow the steps in
Convert DateTime to integer later in this article to create a function that converts
the date/time value in the parameters to match the integer surrogate key of the
source table.
You'll now apply a filter based on conditions in the RangeStart and RangeEnd
parameters.

1. In Power Query Editor, select the date column you want to filter on, and then
choose the dropdown arrow > Date Filters > Custom Filter.

2. In Filter Rows, to specify the first condition, select is after or is after or equal to,
then choose Parameter, and then choose RangeStart.

To specify the second condition, if you selected is after in the first condition, then
choose is before or equal to, or if you selected is after or equal to in the first
condition, then choose is before for the second condition, then choose Parameter,
and then choose RangeEnd.

Important: Verify queries have an equal to (=) on either RangeStart or RangeEnd,

but not both. If the equal to (=) exists on both parameters, a row could satisfy the
conditions for two partitions, which could lead to duplicate data in the model. For
example, = Table.SelectRows(#"Changed Type", each [OrderDate] >= RangeStart
and [OrderDate] <= RangeEnd) could result in duplicate data if there's an

OrderDate that equals both RangeStart and RangeEnd.

Select OK to close.

3. On the Home ribbon in Power Query Editor, select Close & Apply. Power Query
loads data based on the filters defined by the RangeStart and RangeEnd
parameters, and any other filters you've defined.

Power Query loads only data specified between the RangeStart and RangeEnd
parameters. Depending on the amount of data in that period, the table should
load quickly. If it seems slow and process-intensive, it's likely the query isn't
folding.
Define policy
After you've defined RangeStart and RangeEnd parameters, and filtered data based on
those parameters, you'll define an incremental refresh policy. This policy is applied only
after the model is published to the service, and a manual or scheduled refresh operation
is performed.

1. In Data view, right-click a table in the Data pane and select Incremental refresh.

2. In Incremental refresh and real-time data > Select table, verify or select the table.
The default value of the Select table listbox is the table you selected in Data view.

3. Specify required settings:

In Set import and refresh ranges > Incrementally refresh this table move the
slider to On. If the slider is disabled, it means the Power Query expression for the
table doesn't include a filter based on the RangeStart and RangeEnd parameters.

In Archive data starting, specify the historical store period you want to include in
the model. All rows with dates in this period will be loaded into the model in the
service, unless other filters apply.

In Incrementally refresh data starting, specify the refresh period. All rows with
dates in this period will be refreshed in the model each time a manual or
scheduled refresh operation is performed by the Power BI service.
4. Specify optional settings:

In Choose optional settings, select Get the latest data in real time with
DirectQuery (Premium only) to include the latest data changes that occurred at
the data source after the last refresh period. This setting causes the incremental
refresh policy to add a DirectQuery partition to the table.

Select Only refresh complete days to refresh only whole days. If the refresh
operation detects a day isn't complete, rows for that whole day aren't refreshed.
This option is automatically enabled when you select Get the latest data in real
time with DirectQuery (Premium only).

Select Detect data changes to specify a date/time column used to identify and
refresh only the days where the data has changed. A date/time column must exist,
usually for auditing purposes, at the data source. This column should not be the
same column used to partition the data with the RangeStart and RangeEnd
parameters. The maximum value of this column is evaluated for each of the
periods in the incremental range. If it hasn't changed since the last refresh, the
current period isn't refreshed. For models published to Premium capacities, you
can also specify a custom query. To learn more, see Advanced incremental refresh -
Custom queries for detect data changes.

Depending on your settings, your policy should look something like this:
 5. Review your settings and then select Apply to complete the refresh policy. This
step doesn't load data.

Save and publish to the service

Now that your RangeStart and RangeEnd parameters, filtering, and refresh policy
settings are complete, save your model, and then publish to the service. If your model
will become large, be sure to enable Large model storage format before invoking the
first refresh in the service.

Refresh model
In the service, refresh the model. The first refresh will load both new and updated data
in the refresh period as well as historical data for the entire store period. Depending on
the amount of data, this refresh can take quite a while. Subsequent refreshes, whether
manual or scheduled, are typically much faster because the incremental refresh policy is
applied and only data for the period specified in the refresh policy setting is refreshed.

Convert DateTime to integer

This task is only required if your table uses integer surrogate keys instead of Date/Time
values in the date column you use for the RangeStart and RangeEnd filter definition.

The data type of the RangeStart and RangeEnd parameters must be of date/time data
type regardless of the data type of the date column. However, for many data sources,
tables don't have a column of date/time data type but instead have a date column of
integer surrogate keys in the form of yyyymmdd . You typically can't convert these integer
surrogate keys to the Date/Time data type because the result would be a non-folding
query expression, but you can create a function that converts the date/time value in the
parameters to match the integer surrogate key of the data source table without losing
foldability. The function is then called in a filter step. This conversion step is required if
the data source table contains only a surrogate key as integer data type.

1. On the Home ribbon in Power Query Editor, select the New Source dropdown and
then choose Blank Query.

2. In Query Settings, enter a name, for example, DateKey, and then in the formula
editor, enter the following formula:

= (x as datetime) => Date.Year(x)*10000 + Date.Month(x)*100 + Date.Day(x)

3. To test the formula, in Enter Parameter, enter a date/time value, and then select
Invoke. If the formula is correct, an integer value for the date is returned. After
verifying, delete this new Invoked Function query.
 4. In Queries, select the table, and then edit the query formula to call the function
with the RangeStart and RangeEnd parameters.

= Table.SelectRows(#"Reordered Column OrderDateKey", each [OrderDateKey] >

DateKey(RangeStart) and [OrderDateKey] <= DateKey(RangeEnd))

Related content
Troubleshoot configuring incremental refresh
Advanced incremental refresh with the XMLA endpoint
Configure scheduled refresh

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Advanced incremental refresh and real-
time data with the XMLA endpoint
Article • 11/10/2023

Semantic models in a Premium capacity with the XMLA endpoint enabled for read/write
operations allow more advanced refresh, partition management, and metadata only
deployments through tool, scripting, and API support. In addition, refresh operations
through the XMLA endpoint aren't limited to 48 refreshes per day, and the scheduled
refresh time limit isn't imposed.

Partitions
Semantic model table partitions aren't visible and can't be managed by using Power BI
Desktop or the Power BI service. For models in a workspace assigned to a Premium
capacity, partitions can be managed through the XMLA endpoint by using tools like SQL
Server Management Studio (SSMS), the open-source Tabular Editor, scripted with
Tabular Model Scripting Language (TMSL), and programmatically with the Tabular Object
Model (TOM).

When you first publish a model to the Power BI service, each table in the new model has
one partition. For tables with no incremental refresh policy, that one partition contains
all rows of data for that table, unless filters have been applied. For tables with an
incremental refresh policy, that one initial partition only exists because Power BI hasn't
yet applied the policy. You configure the initial partition in Power BI Desktop when you
define the date/time range filter for your table based on the RangeStart and RangeEnd
parameters, and any other filters applied in Power Query Editor. This initial partition
contains only those rows of data that meet your filter criteria.

When you perform the first refresh operation, tables with no incremental refresh policy
refresh all rows contained in that table's default single partition. For tables with an
incremental refresh policy, refresh and historical partitions are automatically created and
rows are loaded into them according to the date/time for each row. If the incremental
refresh policy includes getting data in real time, Power BI also adds a DirectQuery
partition to the table.

This first refresh operation can take quite some time depending on the amount of data
that needs to be loaded from the data source. The complexity of the model can also be
a significant factor because refresh operations must do more processing and
recalculation. This operation can be bootstrapped. For more information, see Prevent
timeouts on initial full refresh.

Partitions are created for and named by period granularity: Years, quarters, months, and
days. The most recent partitions, the refresh partitions, contains rows in the refresh
period you specify in the policy. Historical partitions contain rows by complete period
up to the refresh period. If real time is enabled, a DirectQuery partition picks up any
data changes that occurred after the end date of the refresh period. Granularity for
refresh and historical partitions is dependent on the refresh and historical (store) periods
you choose when defining the policy.

For example, if today's date is February 2, 2021 and our FactInternetSales table at the
data source contains rows up through today, if our policy specifies to include real-time
changes, refresh rows in the last one day refresh period, and store rows in the last three
years historical period. Then with the first refresh operation, a DirectQuery partition is
created for changes in the future, a new import partition is created for today's rows, a
historical partition is created for yesterday, a whole day period, February 1, 2021. A
historical partition is created for the previous whole month period (January 2021), a
historical partition is created for the previous whole year period (2020), and historical
partitions for 2019 and 2018 whole year periods are created. No whole quarter
partitions are created because we haven't yet completed the first full quarter of 2021.

With each refresh operation, only the refresh period partitions are refreshed and the
date filter of the DirectQuery partition is updated to include only those changes that
occur after the current refresh period. A new refresh partition is created for new rows
with a new date/time within the updated refresh period, and existing rows with a
date/time already within existing partitions in the refresh period are refreshed with
updates. Rows with a date/time older than the refresh period are no longer refreshed.

As whole periods close, partitions are merged. For example, if a one-day refresh period
and three year historical store period is specified in the policy, on the first day of the
month, all day partitions for the previous month are merged into a month partition. On
the first day of a new quarter, all three previous month partitions are merged into a
quarter partition. On the first day of a new year, all four previous quarter partitions are
merged into a year partition.
A model always retains partitions for the entire historical store period plus whole period
partitions up through the current refresh period. In the example, a full three years of
historical data are retained in partitions for 2018, 2019, 2020, and also partitions for the
2021Q101 month period, the 2021Q10201 day period, and the current day refresh
period partition. Because the example retains historical data for three years, the 2018
partition is retained until the first refresh on January 1, 2022.

With Power BI incremental refresh and real-time data, the service handles the partition
management for you based on the policy. While the service can handle all of the
partition management for you, by using tools through the XMLA endpoint, you can
selectively refresh partitions individually, sequentially, or in parallel.

Refresh management with SQL Server

Management Studio
SQL Server Management Studio (SSMS) can be used to view and manage partitions
created by the application of incremental refresh policies. By using SSMS you can, for
example, refresh a specific historical partition not in the incremental refresh period to
perform a back-dated update without having to refresh all historical data. SSMS can also
be used when bootstrapping to load historical data for large models by incrementally
adding/refreshing historical partitions in batches.
Override incremental refresh behavior
With SSMS, you also have more control over how to invoke refreshes by using Tabular
Model Scripting Language and the Tabular Object Model. For example, in SSMS, in
Object Explorer, right-click a table and then select the Process Table menu option, and
then select the Script button to generate a TMSL refresh command.
These parameters can be used with the TMSL refresh command to override the default
incremental refresh behavior:

applyRefreshPolicy. If a table has an incremental refresh policy defined,

applyRefreshPolicy determines if the policy is applied or not. If the policy isn't

applied, a process full operation leaves partition definitions unchanged and all
partitions in the table are fully refreshed. Default value is true.

effectiveDate. If an incremental refresh policy is being applied, it needs to know

the current date to determine rolling window ranges for the incremental refresh
and historical periods. The effectiveDate parameter allows you to override the
current date. This parameter is useful for testing, demos, and business scenarios
where data is incrementally refreshed up to a date in the past or the future, for
example, budgets in the future. The default value is the current date.

JSON

{
"refresh": {
"type": "full",

"applyRefreshPolicy": true,
"effectiveDate": "12/31/2013",

"objects": [
{
"database": "IR_AdventureWorks",
"table": "FactInternetSales"
}
]
 }
}

To learn more about overriding default incremental refresh behavior with TMSL, see
Refresh command.

Ensuring optimal performance

With each refresh operation, the Power BI service might send initialization queries to the
data source for each incremental refresh partition. You might be able to improve
incremental refresh performance by reducing the number of initialization queries by
ensuring the following configuration:

The table you configure incremental refresh for should get data from a single data
source. If the table gets data from more than one data source, the number of
queries sent by the service for each refresh operation is multiplied by the number
of data sources, potentially reducing refresh performance. Ensure the query for the
incremental refresh table is for a single data source.
For solutions with both incremental refresh of import partitions and real-time data
with Direct Query, all partitions must query data from a single data source.
If your security requirements allow, set the Data source privacy level setting to
Organizational or Public. By default, the privacy level is Private, however this level
can prevent data from being exchanged with other cloud sources. To set the
privacy level, select the More options menu and then choose Settings > Data
source credentials > Edit credentials > Privacy level setting for this data source.
If Privacy level is set in the Power BI Desktop model before publishing to the
service, it isn't transferred to the service when you publish. You must still set it in
semantic model settings in the service. To learn more, see Privacy levels.
If using an On-premises Data Gateway, be sure you’re using version 3000.77.3 or
higher.

Prevent timeouts on initial full refresh

After you publish to the Power BI service, the initial full refresh operation for the model
creates partitions for the incremental refresh table, loads, and processes historical data
for the entire period defined in the incremental refresh policy. For some models that
load and process large amounts of data, the amount of time the initial refresh operation
takes can exceed the refresh time limit imposed by the service or a query time limit
imposed by the data source.
Bootstrapping the initial refresh operation allows the service to create partition objects
for the incremental refresh table, but not load and process historical data into any of the
partitions. SSMS is then used to selectively process partitions. Depending on the amount
of data to be loaded for each partition, you can process each partition sequentially or in
small batches to reduce the potential for one or more of those partitions to cause a
timeout. The following methods work for any data source.

Apply Refresh Policy

The open-source Tabular Editor 2 tool provides an easy way to bootstrap an initial
refresh operation. After publishing a model with an incremental refresh policy defined
for it from Power BI Desktop to the service, connect to the model by using the XMLA
endpoint in Read/Write mode. Run Apply Refresh Policy on the incremental refresh
table. With only the policy applied, partitions are created but no data is loaded into
them. Then connect with SSMS to refresh the partitions sequentially or in batches to
load and process the data. For more information, see Incremental refresh in the
Tabular editor documentation.

Power Query filter for empty partitions

Prior to publishing the model to the service, in Power Query Editor, add another filter to
the ProductKey column that filters out any value other than 0, effectively or filtering out
all data from the FactInternetSales table.
After selecting Close & Apply in Power Query Editor, defining the incremental refresh
policy, and saving the model, the model is published to the service. From the service, the
initial refresh operation is run on the model. Partitions for the FactInternetSales table
are created according to the policy, but no data is loaded and processed because all
data is filtered out.

After the initial refresh operation is complete, back in Power Query Editor, the other
filter on the ProductKey column is removed. After selecting Close & Apply in Power
Query Editor and saving the model, the model is not published again. If the model is
published again, it overwrites the incremental refresh policy settings and forces a full
refresh on the model when a subsequent refresh operation is performed from the
service. Instead, perform a metadata only deployment by using the Application Lifecycle
Management (ALM) Toolkit that removes the filter on the ProductKey column from the
model. SSMS can then be used to selectively process partitions. When all partitions have
been fully processed, which must include a process recalculation on all partitions, from
SSMS, subsequent refresh operations on the model from the service refresh only the
incremental refresh partitions.

 Tip

Be sure to check out videos, blogs, and more provided by Power BI's community of
BI experts.

Search for "Prevent timeouts with incremental refresh" on Bing .

To learn more about processing tables and partitions from SSMS, see Process database,
table, or partitions (Analysis Services). To learn more about processing models, tables,
and partitions by using TMSL, see Refresh command (TMSL).

Custom queries for detect data changes

TMSL and TOM can be used to override the detected data changes behavior. Not only
can this method be used to avoid persisting the last-update column in the in-memory
cache, it can enable scenarios where a configuration or instruction table is prepared by
extract, transform, and load (ETL) processes for flagging only the partitions that need to
be refreshed. This method can create a more efficient incremental refresh process where
only the required periods are refreshed, no matter how long ago data updates took
place.

The pollingExpression is intended to be a lightweight M expression or name of another

M query. It must return a scalar value and will be executed for each partition. If the value
returned is different to what it was the last time an incremental refresh occurred, the
partition is flagged for full processing.

The following example covers all 120 months in the historical period for backdated
changes. Specifying 120 months instead of 10 years means data compression might not
be quite as efficient, but avoids having to refresh a whole historical year, which would be
more expensive when a month would be sufficient for a backdated change.

JSON

"refreshPolicy": {
"policyType": "basic",
"rollingWindowGranularity": "month",
"rollingWindowPeriods": 120,
"incrementalGranularity": "month",
"incrementalPeriods": 120,
"pollingExpression": "<M expression or name of custom polling query>",
"sourceExpression": [
"let ..."
]
}

 Tip

Be sure to check out videos, blogs, and more provided by Power BI's community of
BI experts.

Search for "Power BI Incremental refresh detect data changes" on Bing .

Metadata only deployment

When publishing a new version of a .pbix file from Power BI Desktop to a workspace, if a
model with the same name already exists, you're prompted to replace the existing
model.

In some cases, you might not want to replace the model, especially with incremental
refresh. The model in Power BI Desktop could be much smaller than the one in the
Power BI service. If the model in the Power BI service has an incremental refresh policy
applied, it might have several years of historical data that will be lost if the model is
replaced. Refreshing all the historical data could take hours and result in system
downtime for users.

Instead, it's better to perform a metadata only deployment, which allows deployment of
new objects without losing the historical data. For example, if you've added a few
measures you can deploy only the new measures without needing to refresh the data,
saving time.

For workspaces assigned to a Premium capacity configured for XMLA endpoint

read/write, compatible tools enable metadata only deployment. For example, the ALM
Toolkit is a schema diff tool for Power BI models and can be used to perform
deployment of metadata only.

Download and install the latest version of the ALM Toolkit from the Analysis Services Git
repo . Step-by-step guidance on using ALM Toolkit isn't included in Microsoft
documentation. ALM Toolkit documentation links and information on supportability are
available on the Help ribbon. To perform a metadata only deployment, perform a
comparison and select the running Power BI Desktop instance as the source, and the
existing model in the Power BI service as the target. Consider the differences displayed
and skip the update of the table with incremental refresh partitions or use the Options
dialog to retain partitions for table updates. Validate the selection to ensure the integrity
of the target model and then update.
Adding an incremental refresh policy and real-
time data programmatically
You can also use the TMSL and TOM to add an incremental refresh policy to an existing
model through the XMLA endpoint.

７ Note

To avoid compatibility issues, make sure you use the latest version of the Analysis
Services client libraries. For example, to work with Hybrid policies, the version must
be 19.27.1.8 or higher.

The process includes of the following steps:

1. Ensure the target model has the required minimum compatibility level. In SSMS,
right-click the [model name] > Properties > Compatibility Level. To increase the
compatibility level, either use a createOrReplace TMSL script or check the following
TOM sample code for an example.

text
 a. Import policy - 1550
b. Hybrid policy - 1565

2. Add the RangeStart and RangeEnd parameters to the model expressions. If

necessary, also add a function to convert Date/Time values to date keys.

3. Define a RefreshPolicy object with the desired archiving (rolling window) and
incremental refresh periods as well as a source expression that filters the target
table based on the RangeStart and RangeEnd parameters. Set the refresh policy
mode to Import or Hybrid depending on your real-time data requirements. Hybrid
causes Power BI to add a DirectQuery partition to the table to fetch the latest
changes from the data source that occurred after the last refresh time.

4. Add the refresh policy to the table and perform a full refresh so that Power BI
partitions the table according to your requirements.

The following code sample demonstrates how to perform the previous steps by using
TOM. If you want to use this sample as is, you must have a copy for the
AdventureWorksDW database and import the FactInternetSales table into a model. The
code sample assumes that the RangeStart and RangeEnd parameters and the DateKey
function don't exist in the model. Just import the FactInternetSales table and publish
the model to a workspace on Power BI Premium. Then update the workspaceUrl so that
the code sample can connect to your model. Update any more code lines as necessary.

C#

using System;
using TOM = Microsoft.AnalysisServices.Tabular;
namespace Hybrid_Tables
{
class Program
{
static string workspaceUrl = "<Enter your Workspace URL here>";
static string databaseName = "AdventureWorks";
static string tableName = "FactInternetSales";
static void Main(string[] args)
{
using (var server = new TOM.Server())
{
// Connect to the dataset.
server.Connect(workspaceUrl);
TOM.Database database =
server.Databases.FindByName(databaseName);
if (database == null)
{
throw new ApplicationException("Database cannot be
found!");
 }
if(database.CompatibilityLevel < 1565)
{
database.CompatibilityLevel = 1565;
database.Update();
}
TOM.Model model = database.Model;
// Add RangeStart, RangeEnd, and DateKey function.
model.Expressions.Add(new TOM.NamedExpression {
Name = "RangeStart",
Kind = TOM.ExpressionKind.M,
Expression = "#datetime(2021, 12, 30, 0, 0, 0) meta
[IsParameterQuery=true, Type=\"DateTime\", IsParameterQueryRequired=true]"
});
model.Expressions.Add(new TOM.NamedExpression
{
Name = "RangeEnd",
Kind = TOM.ExpressionKind.M,
Expression = "#datetime(2021, 12, 31, 0, 0, 0) meta
[IsParameterQuery=true, Type=\"DateTime\", IsParameterQueryRequired=true]"
});
model.Expressions.Add(new TOM.NamedExpression
{
Name = "DateKey",
Kind = TOM.ExpressionKind.M,
Expression =
"let\n" +
" Source = (x as datetime) => Date.Year(x)*10000
+ Date.Month(x)*100 + Date.Day(x)\n" +
"in\n" +
" Source"
});
// Apply a RefreshPolicy with Real-Time to the target table.
TOM.Table salesTable = model.Tables[tableName];
TOM.RefreshPolicy hybridPolicy = new TOM.BasicRefreshPolicy
{
Mode = TOM.RefreshPolicyMode.Hybrid,
IncrementalPeriodsOffset = -1,
RollingWindowPeriods = 1,
RollingWindowGranularity =
TOM.RefreshGranularityType.Year,
IncrementalPeriods = 1,
IncrementalGranularity = TOM.RefreshGranularityType.Day,
SourceExpression =
"let\n" +
" Source =
Sql.Database(\"demopm.database.windows.net\", \"AdventureWorksDW\"),\n" +
" dbo_FactInternetSales =
Source{[Schema=\"dbo\",Item=\"FactInternetSales\"]}[Data],\n" +
" #\"Filtered Rows\" =
Table.SelectRows(dbo_FactInternetSales, each [OrderDateKey] >=
DateKey(RangeStart) and [OrderDateKey] < DateKey(RangeEnd))\n" +
"in\n" +
" #\"Filtered Rows\""
};
 salesTable.RefreshPolicy = hybridPolicy;
model.RequestRefresh(TOM.RefreshType.Full);
model.SaveChanges();
}
Console.WriteLine("{0}{1}", Environment.NewLine, "Press [Enter]
to exit...");
Console.ReadLine();
}
}
}

Next steps
Partitions in tabular models
External tools in Power BI Desktop
Configure scheduled refresh
Troubleshoot incremental refresh and real-time data
Troubleshoot incremental refresh and
real-time data
Article • 04/26/2024

There are two phases when implementing an incremental refresh and real-time data
solution, the first being configuring parameters, filtering, and defining a policy in Power
BI Desktop, and the second being the initial semantic model refresh operation and
subsequent refreshes in the service. This article discusses troubleshooting separately for
each of these phases.

Having partitioned the table in the Power BI service, it's important to keep in mind that
incrementally refreshed tables that are also getting real-time data with DirectQuery are
now operating in hybrid mode, meaning they operate in both import and DirectQuery
mode. Any tables with relationships to such an incrementally refreshed hybrid table
must use Dual mode so that they can be used in import and DirectQuery mode without
performance penalties. Moreover, report visuals might cache results to avoid sending
queries back to the data source, which would prevent the table from picking up the
latest data updates in real time. The final troubleshooting section covers these hybrid-
mode issues.

Before troubleshooting incremental refresh and real-time data, be sure to review

Incremental refresh for models and real-time data and step-by-step information in
Configure incremental refresh and real-time data.

Configuring in Power BI Desktop

Most problems that occur when configuring incremental refresh and real-time data have
to do with query folding. As described in Incremental refresh for models overview -
Supported data sources, your data source must support query folding.

Problem: Loading data takes too long

In Power Query Editor, after selecting Apply, loading data takes an excessive amount of
time and computer resources. There are several potential causes.

Cause: Data type mismatch

This issue can be caused by a data type mismatch where Date/Time is the required data
type for the RangeStart and RangeEnd parameters, but the table date column on which
the filters are applied aren't Date/Time data type, or vice-versa. Both the parameters
data type and the filtered data column must be Date/Time data type and the format
must be the same. If not, the query can't be folded.

Solution: Verify data type

Verify the date/time column for the incremental refresh table is of Date/Time data type.
If your table doesn't contain a column of Date/Time data type, but instead uses an
integer data type, you can create a function that converts the date/time value in the
RangeStart and RangeEnd parameters to match the integer surrogate key of the data

source table. To learn more, see Configure incremental refresh - Convert DateTime to
integer.

Cause: The data source doesn't support query folding

As described in Incremental refresh and real-time data for models - Requirements,
incremental refresh is designed for data sources that support query folding. Make sure
data source queries are being folded in Power BI Desktop before publishing to the
service, where query folding issues can be significantly compounded. This approach is
especially important when including real-time data in an incremental refresh policy
because the real-time DirectQuery partition requires query folding.

Solution: Verify and test queries

In most cases, a warning is shown in the Incremental refresh policy dialog indicating if
the query to be executed against the data source doesn't support query folding.
However, in some cases it might be necessary to further ensure query folding is
possible. If possible, monitor the query being passed to the data source by using a tool
like SQL Profiler. A query with filters based on RangeStart and RangeEnd must be
executed in a single query.

You can also specify a short date/time period in the RangeStart and RangeEnd
parameters that include no more than a few thousand rows. If the load of filtered data
from the data source to the model takes a long time and is process intensive, it likely
means the query isn't being folded.

If you determine the query isn't being folded, refer to Query folding guidance in Power
BI Desktop and Power Query query folding for help with identifying what might be
preventing query folding and how, or if, the data source can even support query folding.
Semantic model refresh in the service
Troubleshooting incremental refresh issues in the service differ depending on the type
of capacity your model has been published to. Semantic models on Premium capacities
support using tools like SQL Server Management Studio (SSMS) to view and selectively
refresh individual partitions. Power BI Pro models on the other hand don't provide tool
access through the XMLA endpoint, so troubleshooting incremental refresh issues might
require a little more trial and error.

Problem: Initial refresh times out

Scheduled refresh for Power BI Pro models on a shared capacity have a time limit of two
hours. This time limit is increased to five hours for models in a Premium capacity. Data
source systems might also impose a query return size limit or query timeout.

Cause: Data source queries aren't being folded

While problems with query folding can usually be determined in Power BI Desktop
before publishing to the service, it's possible that model refresh queries aren't being
folded, leading to excessive refresh times and query mashup engine resource utilization.
This situation happens because a query is created for every partition in the model. If the
queries aren't being folded, and data isn't being filtered at the data source, the engine
then attempts to filter the data.

Solution: Verify query folding

Use a tracing tool at the data source to determine the query being passed for each
partition is a single query that includes a filter based on the RangeStart and RangeEnd
parameters. If not, verify query folding is occurring in the Power BI Desktop model when
loading a small filtered amount of data into the model. If not, get it fixed in the model
first, perform a metadata only update to the model (by using XMLA endpoint), or if a
Power BI Pro model on a shared capacity, delete the incomplete model in the service,
republish, and try an initial refresh operation again.

If you determine queries aren't being folded, refer to Query folding guidance in Power
BI Desktop and Power Query query folding for help with identifying what might be
preventing query folding.

Cause: Data loaded into partitions is too large

Solution: Reduce model size
In many cases, the timeout is caused by the amount of data that must be queried and
loaded into the model partitions exceeds the time limits imposed by the capacity.
Reduce the size or complexity of your model, or consider breaking the model into
smaller pieces.

Solution: Enable Large model storage format

For models published to Premium capacities, if the model grows beyond 1 GB or more,
you can improve refresh operation performance and ensure the model doesn't max out
size limits by enabling Large model storage format before performing the first refresh
operation in the service. To learn more, see Large models in Power BI Premium.

Solution: Bootstrap initial refresh

For models published to Premium capacities, you can bootstrap the initial refresh
operation. Bootstrapping allows the service to create table and partition objects for the
model, but not load and process historical data into any of the partitions. To learn more,
see Advanced incremental refresh - Prevent timeouts on initial full refresh.

Cause: Data source query timeout

Queries can be limited by a default time limit for the data source.

Solution: Override the time limit in the query expression

Many data sources allow overriding time limit in the query expression. To learn more,
see Incremental refresh for models - Time limits.

Problem: Refresh fails because of duplicate values

Cause: Post dates have changed

With a refresh operation, only data that has changed at the data source is refreshed in
the model. As the data is divided by a date, it's recommended that post (transaction)
dates aren't changed.

If a date is changed accidentally, then two issues can occur: Users notice some totals
changed in the historical data (that isn't supposed to happen), or during a refresh an
error is returned indicating a unique value isn't in fact unique. For the latter, this
situation can happen when the table with incremental refresh configured is used in a
1:N relationship with another table as the 1 side and should have unique values. When

the data is changed for a specific ID, that ID then appears in another partition and the
engine detects the value isn't unique.

Solution: Refresh specific partitions

Where there's a business need to change some past data from the dates, a possible
solution is to use SSMS to refresh all partitions from the point where the change is
located up to the current refresh partition, thus keeping the 1 side of the relationship
unique.

Problem: Data is truncated

Cause: Data source query limit has been exceeded

Some data sources, like Azure Data Explorer, Log Analytics, and Application Insights,
have a limit of 64 MB (compressed) on data that can be returned for an external query.
Azure Data Explorer might return an explicit error, but for others like Log Analytics and
Application Insights, the data returned is truncated.

Solution: Specify smaller refresh and store periods

Specify smaller refresh and store periods in the policy. For example, if you specified a
refresh period of one year, and a query error is returned or data returned is truncated,
try a refresh period of 12 months. You want to ensure queries for the current refresh
partition or any historical partitions based on the Refresh and Store periods don't return
more than 64 MB of data.

Problem: Refresh fails because of partition-key conflicts

Cause: Date in the date column at the data source is updated

The filter on the date column is used to dynamically partition the data into period
ranges in the Power BI service. Incremental refresh isn't designed to support cases where
the filtered date column is updated in the source system. An update is interpreted as an
insertion and a deletion, not an actual update. If the deletion occurs in the historical
range and not the incremental range, it isn't picked up, which can cause data refresh
failures due to partition-key conflicts.
Hybrid mode in the service (Preview)
When Power BI applies an incremental refresh policy with real-time data, it turns the
incrementally refreshed table into a hybrid table that operates in both import and
DirectQuery mode. Notice the DirectQuery partition at the end of the following
partitions list of a sample table. The presence of a DirectQuery partition has implications
for related tables and report visuals that query this table.

Problem: Query performance is poor

Cause: Related tables aren't in Dual mode

Hybrid tables operating in both import and DirectQuery mode require any related tables
to operating in Dual mode so that they can act as either cached or not cached,
depending on the context of the query that's submitted to the Power BI model. Dual
mode enables Power BI to reduce the number of limited relationships in the model and
generate efficient data source queries to ensure good performance. Limited
relationships can't be pushed to the data source requiring Power BI to retrieve more
data than necessary. Because Dual tables can act as either DirectQuery or Import tables,
this situation is avoided.

Solution: Convert related tables to Dual mode

When configuring an incremental refresh policy, Power BI Desktop reminds you to
switch any related tables to Dual mode when you select Get the latest data in real time
with DirectQuery (Premium only). In addition, make sure you review all existing table
relationships in Model View.

Tables currently operating in DirectQuery mode, are easily switched to Dual mode. In the
table properties, under Advanced, select Dual from the Storage mode listbox. Tables
currently operating in import mode, however, require manual work. Dual tables have the
same functional constraints as DirectQuery tables. Power BI Desktop therefore can't
convert import tables because they might rely on other functionality not available in
Dual mode. You must manually recreate these tables in DirectQuery mode and then
convert them to Dual mode. To learn more, see Manage storage mode in Power BI
Desktop.

Problem: Report visuals don’t show the latest data

Cause: Power BI caches query results improve performance and

reduce back-end load
By default, Power BI caches query results, so that queries of report visuals can be
processed quickly even if they're based on DirectQuery. Avoiding unnecessary data
source queries improves performance and reduces data source load, but it might also
mean that the latest data changes at the source aren't included in the results.

Solution: Configure automatic page refresh

To keep fetching the latest data changes from the source, configure automatic page
refresh for your reports in the Power BI service. Automatic page refresh can be
performed in fixed intervals, such as five seconds or ten minutes. When that specific
interval is reached, all visuals in that page send an update query to the data source and
update accordingly. Alternatively, you can refresh visuals on a page based on detecting
changes in the data. This approach requires a change detection measure that Power BI
then uses to poll the data source for changes. Change detection is only supported in
workspaces that are part of a Premium capacity. To learn more, see Automatic page
refresh in Power BI.

Related content
Data refresh in Power BI
Advanced incremental refresh with the XMLA endpoint
Incremental refresh for dataflows

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Refresh semantic models created from
local Power BI Desktop files
Article • 11/10/2023

Power BI supports Refresh now and Schedule refresh for semantic models that are
created from imported local Power BI Desktop files. Power BI supports refresh for any of
the following data sources that you connect to or load with Get data and Power Query
Editor.

Power BI gateway (personal mode)

On-premises data gateway (personal mode) supports refresh for the following data
sources:

All online data sources that appear in Power BI Desktop Get data and Power Query
Editor.
All on-premises data sources that appear in Power BI Desktop Get data and Power
Query Editor, except for Hadoop files (HDFS) and Microsoft Exchange.

On-premises data gateway

On-premises data gateway supports refresh for the following data sources:

Analysis Services Tabular

Analysis Services Multidimensional
SQL Server
SAP HANA
Oracle
Teradata
File
Folder
SharePoint list (on-premises)
Web
OData
IBM DB2
MySQL
Sybase
SAP BW
IBM Informix Database
 ODBC

７ Note

A gateway must be installed and running for Power BI to connect to on-premises

data sources and refresh the semantic model.

Refresh in Power BI Desktop vs. Power BI

service
You can do a one-time, manual refresh in Power BI Desktop by selecting Refresh on the
Home tab of the ribbon. When you select Refresh, the data in the file's model refreshes
with updated data from the original data source.

This kind of refresh from within Power BI Desktop is different from manual or scheduled
refresh in the Power BI service. It's important to understand the distinction.

When you import your Power BI Desktop file from a local drive, data and other
information about the model is loaded into a semantic model in the Power BI service.
You base your reports in the Power BI service on the semantic model. You refresh the
data in the Power BI service, not in Power BI Desktop, because you based your reports
on the semantic model in the service. Because the data sources are external, you can
manually refresh the semantic model by using Refresh now, or you can set up a refresh
schedule by using Schedule refresh.

When you refresh the semantic model, Power BI doesn't connect to the file on the local
drive to query for updated data. Power BI uses information in the semantic model to
connect directly to the data sources, query for updated data, and then load the updated
data into the semantic model.

７ Note

Refreshed data in the semantic model doesn't synchronize back to the file on the
local drive.
Scheduled refresh
When you set up a refresh schedule, Power BI connects directly to the data sources by
using the connection information and credentials in the semantic model. Power BI
queries for updated data, then loads the updated data into the semantic model. Any
visualizations in reports and dashboards that are based on that semantic model also
update.

For details on how to set up scheduled refresh, see Configure scheduled refresh.

Troubleshooting
When things go wrong, it's usually because Power BI can't sign into data sources. Make
sure Power BI can sign into your data sources. If the semantic model connects to an on-
premises data source, the gateway might be offline. If the password you use to sign in
to the data source changes, or Power BI gets signed out, try signing into your data
sources again in Data source credentials.

Be sure to set Send refresh failure notifications to Semantic model owner, so you know
right away if a scheduled refresh fails.

Sometimes refreshing data might not go as you expect. This issue often involves a
gateway. For tools and known issues, see the following gateway troubleshooting articles:

Troubleshoot the on-premises data gateway

Troubleshoot the Power BI Gateway - Personal

More questions? Try asking the Power BI Community

Refresh a semantic model stored on
OneDrive or SharePoint Online
Article • 11/10/2023

Importing files from OneDrive or SharePoint Online into the Power BI service is a great
way to make sure your work in Power BI Desktop stays in sync with the Power BI service.

Advantages of storing a Power BI Desktop file

on OneDrive or SharePoint Online
When you store a Power BI Desktop file on OneDrive or SharePoint Online, any data
you’ve loaded into your file’s model is imported into the semantic model. Any reports
you’ve created from the file are loaded into Reports in the Power BI service. Let's say
you make changes to your file on OneDrive or SharePoint Online. These changes can
include adding new measures, changing column names, or editing visualizations. Once
you save the file, Power BI service syncs with those changes too, usually within about an
hour.

You can do a one-time, manual refresh right in Power BI Desktop by selecting Refresh
on the Home ribbon. When you select Refresh, you refresh the file’s model with
updated data from the original data source. This kind of refresh happens entirely from
within the Power BI Desktop application itself. It's different from a manual or scheduled
refresh in Power BI, and it’s important to understand the distinction.

When you import your Power BI Desktop file from OneDrive or SharePoint Online, you
load data and model information into a semantic model in Power BI. After that, you
refresh the semantic model in the Power BI service because that's what your reports are
based on. Because the data sources are external, you can manually refresh the semantic
model by using Refresh now or you can set up a refresh schedule by using Schedule
refresh.
When you refresh the semantic model, Power BI doesn't connect to the file on OneDrive
or SharePoint Online to query for updated data. It uses information in the semantic
model to connect directly to the data sources and query for updated data. Then, it loads
that data into the semantic model. This refreshed data in the semantic model isn't
synchronized back to the file on OneDrive or SharePoint Online.

Automatic versus manual updates of model

information
By default, Power BI updates model information from OneDrive and SharePoint on an
hourly basis. If you want these updates to occur manually, you can disable automatic
OneDrive refresh in the semantic model settings. Open the semantic model settings,
expand the OneDrive refresh section, and set the toggle to Off.
Semantic model owners versus users with write
permission
By default, semantic model owners and semantic model users with write permission can
manually refresh the model information and data in a semantic model by using Refresh
now. As part of a manual refresh, Power BI retrieves the latest model information from
OneDrive or SharePoint and then refreshes the data. The latest model information can
include new and modified data connections and tables added to the files in OneDrive or
SharePoint.

You can restrict the ability to add new data sources to a semantic model in Power BI by
limiting model information updates to semantic model owners. In the semantic model
settings, expand Sync with OneDrive and SharePoint, select Restrict updates, and then
select Apply.

With restricted updates, only semantic model owners can update the model information
in the semantic model with changes made to the version stored in OneDrive and
SharePoint. Semantic model owners must manually refresh semantic models for the
changes to be reflected. If a semantic model user with write permission refreshes the
semantic model, changes from files stored in OneDrive or SharePoint won't be reflected.
If you want semantic model owners and semantic model users with write permission to
have the ability to update the model information, select Automatic updates. Semantic
models in the Power BI service are automatically updated with changes made to the
versions of the semantic models stored in OneDrive and SharePoint.

Existing semantic models will be set to Default updates. Once the setting is changed to
either Restrict updates or Automatic updates, Default updates will no longer be an
option for the semantic model.

New semantic models will be assigned Restricted updates upon creation. The setting
can be changed to Automatic updates if desired, with no option to apply the Default
updates setting.

The difference between Automatic updates and Default updates is that the Default
updates setting is applied to existing semantic models, while the Automatic updates
setting needs to be applied after a new semantic model is created, since new semantic
models default to Restricted updates.

Setting Who can make updates Refresh Availability Default

name type setting

Restrict Semantic model owners Manual Always an option On new

updates only semantic
models

Automatic Semantic model owners Automatic Always an option Never

updates and semantic model users
with write permission

Default Semantic model owners Automatic Once another setting On existing

updates and semantic model users is applied, no longer semantic
with write permission an option models

Enforcing restricted updates

Tenant admins can enforce restricted updates across all semantic models in their
organization by disabling the tenant setting Semantic model owners can choose to
automatically update semantic models from files imported from OneDrive or
SharePoint.
With restricted updates enforced at the tenant level, semantic model owners can no
longer enable automatic updates in the Sync with OneDrive and SharePoint section. An
information block shows the user that an admin has disabled automatic updates for the
organization.

What’s supported?
Power BI supports Refresh and Schedule refresh for semantic models created from
Power BI Desktop files imported from a local drive where you use Get data or Power
Query Editor to connect to and load data from the following data sources.

７ Note
 Onedrive refresh for live connection semantic models is supported. However,
changing the live connection semantic model, from one semantic model to another
in an already published report, is not supported in the OneDrive refresh scenario.

Power BI Gateway - Personal

All online data sources shown in Power BI Desktop’s Get data and Power Query
Editor.
All on-premises data sources shown in Power BI Desktop’s Get data and Power
Query Editor except for Hadoop File (HDFS) and Microsoft Exchange.

On-premises data gateway

On-premises data gateway supports refresh for the following data sources:

Analysis Services Tabular

Analysis Services Multidimensional
SQL Server
SAP HANA
Oracle
Teradata
File
Folder
SharePoint list (on-premises)
Web
OData
IBM DB2
MySQL
Sybase
SAP BW
IBM Informix Database
ODBC

７ Note

A gateway must be installed and running in order for Power BI to connect to on-
premises data sources and refresh the semantic model.
OneDrive or OneDrive for work or school.
What’s the difference?
If you have both a personal OneDrive and OneDrive for work or school, you should keep
any files you want to import into Power BI in OneDrive for work or school. Here’s why:
You likely use two different accounts to sign into them.

When you connect to OneDrive for work or school in Power BI, connection is easy
because your Power BI account is often the same account as your OneDrive for work or
school account. With personal OneDrive, you usually sign in with a different Microsoft
account .

When you sign in with your Microsoft account, be sure to select Keep me signed in.
Power BI can then synchronize any updates you make in the file in Power BI Desktop
with semantic models in Power BI.

If you've changed your Microsoft credentials, you can't synchronize changes between
your file on OneDrive and the semantic model in Power BI. You need to connect to and
import your file again from OneDrive.

How do I schedule refresh?

When you set up a refresh schedule, Power BI connects directly to the data sources.
Power BI uses connection information and credentials in the semantic model to query
for updated data. Then Power BI loads the updated data into the semantic model. It
then updates any report visualizations and dashboards based on that semantic model in
the Power BI service.

For details on how to set up schedule refresh, see Configure scheduled refresh.

When things go wrong

When things go wrong, it’s usually because Power BI can’t sign into data sources. Things
may also go wrong if the semantic model tries to connect to an on-premises data
source but the gateway is offline. To avoid these issues, make sure Power BI can sign
into data sources. Try signing into your data sources in Data Source Credentials.
Sometimes the password you use to sign into a data source changes or Power BI gets
signed out from a data source.

When you save your changes to the Power BI Desktop file on OneDrive and you don't
see those changes in Power BI within an hour or so, it could be because Power BI can't
connect to your OneDrive. Try connecting to the file on OneDrive again. If you’re
prompted to sign in, make sure you select Keep me signed in. Because Power BI wasn't
able to connect to your OneDrive to synchronize with the file, you need to import your
file again.

Semantic models stored on OneDrive or SharePoint are set to restrict updates by

default. If the semantic model is set to restrict updates, then updates can only happen
when the semantic model owner manually refreshes the semantic model, which can
cause changes to Power BI files on OneDrive and SharePoint to not be reflected in the
Power BI service. A semantic model owner might run into an error message after
updating a file in OneDrive or SharePoint. The semantic model owner can fix the error
by choosing to always manually refresh the semantic model, or changing the semantic
model setting to automatic updates.

If the semantic model owner is unable to change the setting of the semantic model to
automatic updates, the tenant admin has likely enforced restricted updates across all
semantic models in the organization. To enable the semantic model owner to change
the setting, they must contact their Fabric admin and request that the admin enable the
Semantic model owners can choose to automatically update from files imported from
OneDrive or SharePoint setting.

If the semantic model owner has set up scheduled refresh on semantic models, then the
model will still refresh on schedule. However, the other contents of the report, such as
visuals, will not refresh unless manual updates are made.

Import of sensitivity-labeled .pbix files (both protected and unprotected) stored on

OneDrive or SharePoint Online, as well as on-demand and automatic semantic
model refresh from such files, is supported, except for the following scenarios:
Protected live-connected .pbix files and protected Azure Analysis Services .pbix
files. Refresh will fail. Neither report content nor label will be updated.
Labeled unprotected Live Connect .pbix files: Report content will be updated but
label won't be updated.
 When the .pbix file has had a new sensitivity label applied that the semantic
model owner doesn't have usage rights to. In this case, refresh will fail. Neither
report content nor label will be updated.
If the semantic model owner's access token for OneDrive/SharePoint has
expired. In this case, refresh will fail. Neither report content nor label will be
updated.

Troubleshooting
Sometimes refreshing data may not go as expected. You'll typically run into data refresh
issues when you're connected with a gateway. Take a look at the gateway
troubleshooting articles for tools and known issues.

Troubleshooting the On-premises data gateway

Troubleshooting the Power BI Gateway - Personal

More questions? Try asking the Power BI Community .

Refresh a dataset created from an Excel
workbook on a local drive
Article • 08/28/2023

What's supported?

） Important

The following capabilities are deprecated and will no longer be available starting
September 29th, 2023:

Upload of local workbooks to Power BI workspaces will no longer be allowed.

Configuring scheduling of refresh and refresh now for Excel files that don’t
already have scheduled refresh configured will no longer be allowed.

The following capabilities are deprecated and will no longer be available starting
October 31, 2023:

Scheduled refresh and refresh now for existing Excel files that were previously
configured for scheduled refresh will no longer be allowed.
Local workbooks uploaded to Power BI workspaces will no longer open in
Power BI.

After October 31, 2023:

You can download existing local workbooks from your Power BI workspace.
You can publish your Excel data model as a Power BI dataset and schedule
refresh.
You can import Excel workbooks from OneDrive and SharePoint Document
libraries to view them in Power BI.

If your organization uses these capabilities, see more details in Migrating your
Excel workbooks.

In Power BI, Refresh Now and Schedule Refresh is supported for datasets created from
Excel workbooks imported from a local drive where Power Query or Power Pivot is used
to connect to any of the following data sources and load data into the Excel data model.
Power Query is Get & Transform data in Excel 2016.
Power BI Gateway - Personal
All online data sources shown in Power Query.
All on-premises data sources shown in Power Query except for Hadoop file (HDFS)
and Microsoft Exchange.
All online data sources shown in Power Pivot.
All on-premises data sources shown in Power Pivot except for Hadoop file (HDFS)
and Microsoft Exchange.

On-premises data gateway

On-premises data gateway supports refresh for the following data sources:

Analysis Services Tabular

Analysis Services Multidimensional
SQL Server
SAP HANA
Oracle
Teradata
File
Folder
SharePoint list (on-premises)
Web
OData
IBM DB2
MySQL
Sybase
SAP BW
IBM Informix Database
ODBC

Keep the following notes in mind:

A gateway must be installed and running in order for the Power BI service to
connect to on-premises data sources and refresh the dataset.

When using Excel 2013, make sure you've updated Power Query to the latest
version.

Refresh isn't supported for Excel workbooks imported from a local drive where
data exists only in worksheets or linked tables. Refresh is supported for worksheet
 data if it's stored and imported from OneDrive. To learn more, see Refresh a
dataset created from an Excel workbook on OneDrive, or SharePoint Online.

When you refresh a dataset created from an Excel workbook imported from a local
drive, only the data queried from data sources is refreshed.

If you change the structure of the data model in Excel or Power Pivot, for example,
create a new measure or change the name of a column, those changes aren't
copied to the dataset. If you make such changes, reupload or republish the
workbook.

If you expect to make regular changes to the structure of your workbook and you
want those changes to be reflected in the dataset in the Power BI service without
having to reupload, consider putting your workbook on OneDrive. The Power BI
service automatically refreshes both the structure and worksheet data from
workbooks stored and imported from OneDrive.

How do I make sure data is loaded to the Excel

data model?
When you use Power Query to connect to a data source, you have several options where
to load the data. Power Query is Get & Transform data in Excel 2016. To make sure you
load data into the data model, you must select the Add this data to the Data Model
option in the Load To dialog.

７ Note

The images here show Excel 2016.

In Navigator, select Load To…

Or, if you select Edit in Navigator, you open the Query Editor. There you can select Close
& Load To….
Then in Load To, make sure you select Add this data to the Data Model.

What if I use Get External Data in Power Pivot?

No problem. Whenever you use Power Pivot to connect to and query data from an on-
premises or online data source, the data is automatically loaded to the data model.

How do I schedule refresh?

When you set up a refresh schedule, Power BI connects directly to the data sources
using connection information and credentials in the dataset to query for updated data,
then loads the updated data into the dataset. Any visualizations in reports and
dashboards based on that dataset in the Power BI service are also updated.

For details on how to setup schedule refresh, see Configure Schedule Refresh.

When things go wrong

When things go wrong, it's usually because Power BI can't sign into data sources, or if
the dataset connects to an on-premises data source, the gateway is offline. Make sure
Power BI can sign into data sources. If a password you use to sign into a data source
changes, or Power BI gets signed out from a data source, be sure to try signing into your
data sources again in Data Source Credentials.

Be sure to leave the Send refresh failure notification email to me selected. You want to
know right away if a scheduled refresh fails.

） Important

Refresh isn't supported for OData feeds connected to and queried from Power
Pivot. When using an OData feed as a data source, use Power Query.

Troubleshooting
Sometimes refreshing data might not go as expected. Typically problems are caused by
an issue connected with a gateway. Take a look at the gateway troubleshooting articles
for tools and known issues.

Troubleshooting the On-premises data gateway

Troubleshooting the Power BI Gateway - Personal

Next steps
More questions? Try the Power BI Community
Refresh a semantic model created from
an Excel workbook on OneDrive or
SharePoint Online
Article • 11/10/2023

） Important

The following capabilities are deprecated and will no longer be available starting
September 29th, 2023:

Upload of local workbooks to Power BI workspaces will no longer be allowed.

Configuring scheduling of refresh and refresh now for Excel files that don’t
already have scheduled refresh configured will no longer be allowed.

The following capabilities are deprecated and will no longer be available starting
October 31, 2023:

Scheduled refresh and refresh now for existing Excel files that were previously
configured for scheduled refresh will no longer be allowed.
Local workbooks uploaded to Power BI workspaces will no longer open in
Power BI.

After October 31, 2023:

You can download existing local workbooks from your Power BI workspace.
You can publish your Excel data model as a Power BI semantic model and
schedule refresh.
You can import Excel workbooks from OneDrive and SharePoint Document
libraries to view them in Power BI.

If your organization uses these capabilities, see more details in Migrating your
Excel workbooks.

You can import Excel workbooks from your local machine, or from cloud storage such as
OneDrive for work or school or SharePoint Online. This article explores the advantages
of using cloud storage for your Excel files. For more information about how to import
Excel files into Power BI, see Get data from Excel workbook files.
What are the advantages?
When you import files from OneDrive, or SharePoint Online, it ensures the work you’re
doing in Excel stays in sync with the Power BI service. Any data that you’ve loaded into
your file’s model then updates in the semantic model. Any reports you’ve created in the
file load into Reports in Power BI. If you make and save changes to your file on OneDrive
or SharePoint Online, Power BI shows the updates to those changes. For example, if you
add new measures, change column names, or edit visualizations, Power BI reflects the
changes. Your changes typically update within an hour after you've saved them.

When you import an Excel workbook from your personal OneDrive, any data in the
workbook loads into a new semantic model in Power BI. For example, tables in
worksheets, data loaded into the Excel data model, and the structure of the data model
goes into a new semantic model. Power BI automatically connects to the workbook on
OneDrive, or SharePoint Online, approximately every hour to check for updates. If the
workbook changed, Power BI refreshes the semantic model and reports in the Power BI
service.

You can refresh the semantic model in the Power BI service. When you manually refresh
or schedule a refresh on the semantic model, Power BI connects directly to the external
data sources to query for any updated data. It then loads updated data into the
semantic model. Refreshing a semantic model from within Power BI doesn't refresh the
data in the workbook on OneDrive or SharePoint Online.

What’s supported?
Power BI supports the Refresh Now and Schedule Refresh options for semantic models
that meet the following conditions:

The semantic models are created from Power BI Desktop files that are imported
from a local drive.
Get data or Power Query Editor in Power BI is used to connect to and load the
data.
The data is from a source that's described in one of the following sections.

Power BI gateway - personal

All online data sources shown in Power BI Desktop’s Get data and Power Query
Editor.
All on-premises data sources shown in Power BI Desktop’s Get data and Power
Query Editor except for Hadoop file (HDFS) and Microsoft Exchange.
On-premises data gateway
On-premises data gateway supports refresh for the following data sources:

Analysis Services Tabular

Analysis Services Multidimensional
SQL Server
SAP HANA
Oracle
Teradata
File
Folder
SharePoint list (on-premises)
Web
OData
IBM DB2
MySQL
Sybase
SAP BW
IBM Informix Database
ODBC

７ Note

A gateway must be installed and running in order for Power BI to connect to on-
premises data sources and refresh the semantic model.

OneDrive or OneDrive for work or school.

What’s the difference?
If you have both a personal OneDrive and OneDrive for work or school, it’s
recommended you keep files you want to import in OneDrive for work or school. Here’s
why: You likely use two different accounts to sign in and access your files.

In Power BI, connecting to OneDrive for work or school is typically seamless because
you likely use the same account to sign in to Power BI as OneDrive for work or school.
But with personal OneDrive, it's more common to sign in with a different Microsoft
account .
When you sign in to OneDrive for work or school with your Microsoft account, select
Keep me signed in. Power BI can then synchronize any updates you make in the file in
Power BI Desktop with semantic models in Power BI.

If your Microsoft account credentials change, edits to your file on OneDrive can't
synchronize with the semantic model or reports in Power BI. You need to reconnect and
import the file again from your personal OneDrive.

Options for connecting to an Excel file

When you connect to an Excel workbook in OneDrive for work or school, or SharePoint
Online, you have two options on how to get what’s in your workbook into Power BI.

Import Excel data into Power BI – When you import an Excel workbook from your
OneDrive for work or school or SharePoint Online, it works as described previously.

Connect, manage, and view Excel in Power BI – When using this option, you create a
connection from Power BI right to your workbook on OneDrive for work or school or
SharePoint Online.

When you connect to an Excel workbook this way, a semantic model isn't created in
Power BI. But the workbook appears in the Power BI service under Reports with an Excel
icon next to the name. Unlike with Excel Online, when you connect to your workbook
from Power BI, if your workbook has connections to external data sources that load data
into the Excel data model, you can set up a refresh schedule.

When you set up a refresh schedule this way, the only difference is refreshed data goes
into the workbook’s data model on OneDrive, or SharePoint Online, rather than a
semantic model in Power BI.

How do I make sure data is loaded to the Excel

data model?
When you use Power Query (Get & Transform Data in Excel 2016) to connect to a data
source, you have several options of where to load the data. To ensure that you load data
into the data model, you must select the Add this data to the Data Model option in the
Import Data dialog box.

1. In Excel, select Data > Get Data and select where you want your data to come
from. In this example, the data loads from an Excel workbook file.

2. In the file browser window, locate and select your data file and then select Import.

3. In Navigator, select your file and choose Load To… .

Or, in Excel, select Data > Get Data > Launch Power Query Editor to open the
Query Editor. There you can select Close & Load To….

4. Then, in Import Data, be sure to select Add this data to the Data Model and
select OK.
What if I use Get External Data in Power Pivot?
No problem. Whenever you use Power Pivot to connect to and query data from an on-
premises or online data source, the data automatically loads to the data model.

How do I schedule a refresh?

When you set up a refresh schedule, Power BI connects directly to the data sources by
using connection information and credentials in the semantic model to query for
updated data. It then loads the updated data into the semantic model. Any
visualizations in reports and dashboards based on that semantic model in the Power BI
service also update.

For more information about how to set up a scheduled refresh, see Configure scheduled
refresh.

When things go wrong

When things go wrong, it’s usually because Power BI can’t sign in to data sources. Or it's
because the semantic model connects to an on-premises data source and the gateway
is offline. Be sure Power BI can sign in to data sources. If a password you use to sign in
to a data source changes, or Power BI is signed out from a data source, be sure to sign
in to your data sources again in Data Source Credentials.

Be sure to leave the Send refresh failure notification email to me setting selected. You
want to know right away if a scheduled refresh fails.
Important notes
Refresh isn't supported for OData feeds connected to and queried from Power Pivot.
When using an OData feed as a data source, use Power Query.

Troubleshooting
Sometimes refreshing data might not go as expected. Typically, problems with
refreshing are an issue with the data gateway. For tools, tips, and known issues, see the
following articles about troubleshooting the gateway.

Troubleshoot the on-premises data gateway

Troubleshoot the Power BI gateway - personal

More questions? Try the Power BI Community .

Refresh a semantic model created from
a .CSV file on OneDrive or SharePoint
Article • 11/10/2023

When you connect to a comma separated value (.csv) file on OneDrive or SharePoint, a
semantic model is created in Power BI. Data from the .csv file is imported into the
semantic model in Power BI. Power BI then automatically connects to the file and
refreshes any changes with the semantic model in Power BI. If you edit the .csv file in
OneDrive, or SharePoint, after you save, those changes will appear in Power BI, usually
within about an hour. Any visualizations in Power BI based on the semantic model are
automatically updated.

７ Note

By default, using the Get Data experience for specific file type connectors in Power
BI Desktop uses a local reference to the file stored on OneDrive, which will not
automatically update unless you have a gateway configured. To have your CSV
automatically update without having to configure a gateway, rather than using the
Text/CSV connector, use the Web connector and reference the online version of
your CSV.

Advantages
If your files are in a shared folder on OneDrive for work or school, or SharePoint, other
users can work on the same file. After they save the file, changes are automatically
updated in Power BI, usually within an hour.

Many organizations run processes that automatically query databases for data that's
saved to a .csv file each day. If the file is stored on OneDrive, or SharePoint, and the
same file is overwritten each day, as opposed to a new file with a different name being
created each day, you can connect to that file in Power BI. Your semantic model that
connects to the file will be synchronized soon after the file on OneDrive, or SharePoint,
is updated. Any visualizations based on the semantic model are automatically updated.

What’s supported
Comma separated value files are simple text files, connections to external data sources
and reports aren't supported. You can't schedule a refresh on a semantic model created
from a .csv file. However, when the file is on OneDrive, or SharePoint, Power BI will
automatically synchronize any changes to the file with the semantic model about every
hour.

What's the difference between personal

OneDrive and OneDrive for work or school
If you have both a personal OneDrive and OneDrive for work or school, we
recommended you keep any files you want to connect to Power BI on OneDrive for work
or school. Why? Because you likely use two different accounts to sign into them.

Connecting to OneDrive for work or school and Power BI is typically seamless because
the same account you use to sign into Power BI is often the same account used to sign
into OneDrive for work or school. But, with personal OneDrive, you're likely to use a
different Microsoft account .

When you sign into your Microsoft account, be sure to select Keep me signed in. Power
BI can then synchronize any updates with semantic models in Power BI.

When you make changes to your .csv file on OneDrive, and they don't synchronize with
the semantic model in Power BI. It might be because your Microsoft account credentials
changed, you’ll need to connect to the file and import it again from your personal
OneDrive.

When things go wrong

If data in the .csv file on OneDrive changes, and the changes aren’t reflected in Power BI,
it’s most likely because Power BI can't connect to OneDrive. Try connecting to the file
and importing it again. If you’re prompted to sign in, make sure you select Keep me
signed in.

Next steps
Troubleshoot gateways - Power BI
Troubleshoot refresh scenarios

More questions? Ask the Power BI Community

Query caching in Power BI Premium or
Power BI Embedded
Article • 05/21/2024

Organizations with Power BI Premium or Power BI Embedded can take advantage of

query caching to speed up reports associated with a semantic model. Query caching
instructs the Power BI Premium or Power BI Embedded capacity to use its local caching
service to maintain query results, avoiding having the underlying data source compute
those results.

） Important

Query caching is only available on Power BI Premium or Power BI Embedded, for

Import semantic models. It is not applicable DirectQuery or LiveConnect semantic
models that use Azure Analysis Services or SQL Server Analysis Services.

The caching is performed the first time a user opens the report. The service only does
query caching for the initial page that they land on. In other words, queries aren't
cached when you interact with the report. Cached query results are specific to user and
semantic model context and always respect security rules. The query cache respects
personal bookmarks and persistent filters , so queries generated by a personalized
report are cached. Dashboard tiles that are powered by the same queries also benefit
once the query is cached. Performance especially benefits when a semantic model is
accessed frequently and doesn't need to be refreshed often. Query caching can also
reduce load on your capacity by reducing the overall number of queries.

You control query caching behavior on the Settings page for the semantic model in the
Power BI service. It has three possible settings:

Capacity default: Query caching Off

Off: Don't use query caching for this semantic model.
On: Use query caching for this semantic model.
Considerations and limitations
When you change caching settings from On to Off, all previously saved query
results for the semantic model are removed from the capacity cache. You can turn
off caching either explicitly or by reverting to capacity default setting that an
administrator sets to Off. Turning it off can introduce a small delay the next time
any report runs queries against this semantic model. The delay is caused by those
report queries running on demand and not applying saved results. Also, the
required semantic model might need to be loaded into memory before it can
service queries.
The query cache is refreshed when Power BI performs a semantic model refresh.
When the query cache is refreshed, Power BI must run queries against the
underlying data models to get the latest results. If a large number of semantic
models have query caching enabled and the Premium/Embedded capacity is under
heavy load, some performance degradation might occur during cache refresh.
Degradation results from the increased volume of queries being executed.

Related content
What is Power BI embedded analytics?

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

What are Power BI template apps?
Article • 11/10/2023

The new Power BI template apps enable Power BI partners to build Power BI apps with
little or no coding, and deploy them to any Power BI customer. This article is an
overview of the Power BI template app program.

As a Power BI partner, you create a set of out-of-the-box content for your customers
and publish it yourself.

You build template apps that allow your customers to connect and instantiate within
their own accounts. As domain experts, they can unlock the data in a way that's easy for
their business users to consume.

You submit a template app to the Partner center. The apps then become publicly
available in the Power BI apps marketplace and on the Microsoft commercial
marketplace . Here's a high-level look at the public template app creation experience.

Power BI Apps marketplace

Power BI template apps allow Power BI Pro or Power BI Premium users to gain
immediate insights through prepackaged dashboards and reports that can be
connected to live data sources. Many Power BI Apps are already available in the Power
BI Apps marketplace.
 ７ Note

Marketplace apps aren't available for US government cloud instances. For more
information, see Power BI for US government customers.

Process
The general process to develop and submit a template app involves several stages.
Some stages can include more than one activity at the same time.

Stage Power BI Power BI service Partner Center

Desktop

One Build a data Create a workspace. Import .pbix Register as a partner

model and file. Create a complementary
report in a .pbix dashboard
file

Two Create a test package and run

internal validation

Three Promote the test package to With your preproduction

preproduction for validation package, create a Power BI
outside your Power BI tenant, and template app offer and start the
submit it to AppSource validation process

Four Promote the preproduction Go live

package to production

Before you begin

To create the template app, you need permissions to create one. For more information,
see Template app tenant settings.

To publish a template app to the Power BI service and AppSource, you must meet the
requirements for becoming a Partner Center publisher.

High-level steps
Here are the high-level steps.

1. Review the requirements to make sure you meet them.

2. Build a report in Power BI Desktop. Use parameters so you can save it as a file
other people can use.

3. Create a workspace for your template app in your tenant on the Power BI service
( app.powerbi.com ).

4. Import your .pbix file and add content such as a dashboard to your app.

5. Create a test package to test the template app yourself within your organization.
 6. Promote the test app to pre-production to submit the app for validation in
AppSource, and to test outside your own tenant.

7. Submit the content to Partner center for publishing.

8. Make your offer go Live in AppSource, and move your app to production in Power
BI.

9. Now you can start developing the next version in the same workspace, in
preproduction.

Requirements
To create the template app, you need permissions to create one. For more information,
see Template app tenant settings.

To publish a template app to the Power BI service and AppSource, you must meet the
requirements for becoming a Partner Center publisher.

７ Note

Template apps submissions are managed in Partner Center. Use the same
Microsoft Developer Center registration account to sign in. You should have only
one Microsoft account for your AppSource offerings. Accounts shouldn't be specific
to individual services or offers.

Tips
Make sure your app includes sample data to get everyone started in a click.
Limit semantic model size (rule of thumb: .pbix file < 10MBs). This typically means
keeping the size of sample data as small as possible.
Carefully examine your application by installing it in your tenant and in a
secondary tenant. Make sure customers only see what you want them to see.
Use AppSource as your online store to host your application. This way everyone
using Power BI can find your app.
Consider offering more than one template app for separate unique scenarios.
Enable data customization. For example, support custom connection and
parameters configuration by the installer.
If you're an independent software vendor and are distributing your app through
your web service, consider automating parameter configuration during installation
to make things easier for your customers and to increase the likelihood of a
 successful installation. For more information, see Automated configuration of a
template app installation.

See Tips for authoring template apps in Power BI for more suggestions.

Known limitations
Feature Known Limitation

Contents: Exactly one semantic model should be present. Only semantic models built
Semantic models into Power BI Desktop (.pbix files) are allowed. Not supported: Semantic
models from other template apps, cross-workspace semantic models,
paginated reports (.rdl files), and Excel workbooks.

Contents: Reports A single template app can't include more than 20 reports.

Contents: Real-time tiles aren't allowed. In other words, no support for push or
Dashboards streaming datasets.

Contents: Not supported: Dataflows.

Dataflows

Contents from Only .pbix files are allowed. Not supported: .rdl files (paginated reports) and
files Excel workbooks.

Data sources Data sources supported for cloud Scheduled Data refresh are allowed. Not
supported: Live connections, on-premises data sources (personal and
enterprise gateways aren't supported), real time (no support for push
dataset), and composite models.

Semantic model: No cross-workspace semantic models are allowed.

cross-workspace

Query parameters Not supported: Parameters of type Any, Date, or Binary type block refresh
operation for semantic model.

Incremental Template apps don't support incremental refresh.

refresh

Power BI visuals Only publicly available Power BI visuals are supported. Organizational Power
BI visuals aren't supported.

Sovereign clouds Template apps aren't available in sovereign clouds.

Composite Composite models shouldn't be used in the app builder workspace. App
models installers can use composite models after installing the app.

Large semantic Large semantic model storage format isn't supported for template apps.
model storage
format
Support
For support during development, use https://powerbi.microsoft.com/support . We
actively monitor and manage this site. Customer incidents quickly find their way to the
appropriate team.

Next steps
Create a template app
Create a template app in Power BI
Article • 03/21/2023

This article contains step-by-step instructions for creating a Power BI template app.
Power BI template apps let Power BI partners build Power BI apps with little or no
coding, and deploy them to any Power BI customer.

If you can create Power BI reports and dashboards, you can become a template app
builder and build and package analytical content into an app. You can then deploy your
app to other Power BI tenants through any available platform, such as AppSource or
your own web service. If you're distributing your template app through your own web
service, you can automate part of the installation process to make things easier for your
customers.

Power BI admins govern and control who in their organization can create template apps,
and who can install them. Authorized users can install your template app, modify it, and
distribute it to the Power BI consumers in their organizations.

Prerequisites
Here are the requirements for building a template app:

A Power BI pro license

Power BI Desktop (optional)
Familiarity with basic Power BI concepts
Permissions to share a template app publicly as shown in Template app tenant
settings

Create the template workspace

To create a template app you can distribute to other Power BI tenants, you need to
create it in a workspace.

1. In the Power BI service, create a workspace as described in Create a workspace in

Power BI. In the Advanced section, select Develop a template app.
 ） Important

The capacity that the app builder workspace is assigned to does not
determine the capacity assignment of workspaces where app installers install
the app. This means that an app developed in a premium capacity workspace
will not necessarily be installed on a premium capacity workspace. Therefore it
is not recommended to use premium capacity for the builder workspace, as
installer workspaces might not be premium capacity, and functionality that
relies on premium capacity won't work unless the installer manually reassigns
the installed workspace to premium capacity.

2. When you're done creating the workspace, select Save.

７ Note

You need permissions from your Power BI admin to promote template apps.

Add content to the template app workspace

As with a regular Power BI workspace, your next step is to add content to the
workspace. If you're using parameters in Power Query, make sure they have well-defined
types, such as Text . The types Any and Binary aren't supported.

For suggestions to consider when creating reports and dashboards for your template
app, see Tips for authoring template apps in Power BI.

Define the properties of the template app

Now that you have content in your workspace, you can package it in a template app.
The first step is to create a test template app, accessible only from within your
organization on your tenant.

1. In the template app workspace, select Create app.

Next, fill in more options for your template app in six tabs.

2. On the Branding tab, complete the following fields:

App name
Description
Support site. The support link appears under app info after you redistribute
the template app as an organizational app.
App logo. The logo has a 45K file-size limit, must have a 1:1 aspect ratio, and
must be in a .png, .jpg, or .jpeg file format.
App theme color
3. On the Navigation tab, you can turn on New navigation builder to define the
navigation pane of the app.

If you don't turn on New navigation builder, you have the option of selecting an
app landing page. Define a report or dashboard to be the landing page of your
app. Use a landing page that gives the impression you want.
4. On the Control tab, set your app users' limits and restrictions on your app's
content. You can use this control to protect intellectual property in your app.

７ Note

If you want to protect your data, disable the Download the report to file
option and then configure the other two options as desired.

Why:

The view, edit, and export controls on this tab apply only to the Power BI
service. Once you download the .pbix file, it is no longer in the service. It puts
a copy of your data, unprotected, in a location chosen by the user. You then
no longer have any control over what the user can do with it.

If you want to limit access to your queries and measures while still allowing
your users to add their own data sources, consider checking only the Export
or externally connect to data options. This enables users to add their own
data sources without being able to edit your dataset. For more information,
see Use composite models in Power BI Desktop.

5. Parameters are created in the original .pbix file (learn more about creating query
parameters ). You use the capabilities on this tab to help the app installer
configure the app after installation when they connect to their data.
Each parameter has a name, which comes from the query, and a Value field. There
are three options for getting a value for the parameter during installation:

You can require the user who installs the app to enter a value.

In this case, you provide an example that the user replaces. To configure a
parameter in this way, select the Required checkbox, and then give an
example in the textbox that shows the user what kind of value is expected, as
shown in the following example.

You can provide a pre-populated value that the user who installs the app
can't change.

A parameter configured in this way is hidden from the user who installs the
app. You should use this method only if you're sure that the pre-populated
 value is valid for all users. If not, use the first method that requires user input.

To configure a parameter in this way, enter the value in the Value textbox,
and then select the lock icon so the value can't be changed. The following
example shows this option:

You can provide a default value that the user can change during installation.

To configure a parameter in this way, enter the desired default value in the
Value textbox, and leave the lock icon unlocked, as in the following example:

In this tab, you also provide a link to the app documentation.

6. On the Authentication tab, select the authentication method to use. The available
options depend on the data source types being used.
 Privacy level is configured automatically:

A single datasource is automatically configured as private.

A multi anonymous datasource is automatically configured as public.

7. In the test phase, on the Access tab decide who else in your organization can
install and test your app. You will come back and change these settings later. The
setting doesn't affect access of the distributed template app.

8. Select Create app.

You see a message that the test app is ready, with a link to copy and share with
your app testers.

You've also done the first step of the following release management process.

Manage the template app release

Before you release the template app publicly, you want to make sure it's ready. In the
Power BI release management pane, you can follow and inspect the full app release
path. You can also trigger the transition from stage to stage. The common stages are:

Generate a test app for testing within your organization only.

Promote the test package to pre-production stage and test outside of your
organization.
Promote the pre-production package to the production version in Production.
Delete any package or start over from a previous stage.

The URL doesn't change as you move between release stages. Promotion doesn't affect
the URL itself.

To go through the release stages:

1. In the template workspace, select Release Management.

2. If you followed the steps in this article to create the test app, the dot next to
Testing will already be filled in. Select Get link.

If you haven't created the app yet, select Create app to start the template app
creation process.
3. To test the app installation experience, copy the link in the window and paste it
into a new browser window.

From here, you follow the same procedure your app installers will follow. For more
information, see Install and distribute template apps in your organization.

4. In the dialog box, select Install.

5. After installation succeeds, select the app in the Apps list to open it.

6. Verify that the test app has the sample data. To make any changes, go back to the
app in the original workspace. Update the test app until you're satisfied.

7. When you're ready to promote your app to pre-production for testing outside
your tenant, go back to the Release Management pane and select Promote app.
 ７ Note

When you promote the app, it becomes publicly available outside your
organization.

If you don't see the Promote app option, contact your Power BI admin to grant
you permissions for template app development in the admin portal.

8. In the dialog box, select Promote.

9. Copy the new URL to share outside your tenant for testing. This link is also the one
you submit to begin the process of distributing your app on AppSource by
creating a new Partner center offer.
 Submit only pre-production links to the Partner center. After the app is approved
and you get notification that it's published in AppSource, you can promote the
package to production in Power BI.

10. When your app is ready for production or sharing via AppSource, go back to the
Release Management pane and select Promote app next to Pre-production.

11. Select Promote.

Now your app is in production and ready for distribution.

To make your app widely available to Power BI users throughout the world, submit it to
AppSource. For more information, see the Create a Power BI app offer.

Automate parameter configuration during

installation
If you're an independent software vendor and distribute your template app via your web
service, you can create automation that configures template app parameters
automatically when your customers install the app in Power BI. Automatic configuration
makes things easier for your customers and increases the likelihood of a successful
installation, because customers don't have to supply details that they might not know.
For more information, see Automated configuration of a template app installation.

Next steps
To learn how your customers interact with your template app, see Install,
customize, and distribute template apps in your organization.
For details on distributing your app, see the Create a Power BI app offer.
Tips for authoring template apps in
Power BI
Article • 02/06/2023

When you're authoring your template app in Power BI, part of the process is the
logistics of creating the workspace, testing it, and production. But the other important
part is obviously authoring the report and the dashboard. You can break down the
authoring process into several components. Working on these components helps you
create the best possible template app:

Queries. With queries, you connect and transform the data, and define
parameters .
Data model. In the data model, you create relationships, measures, and Q&A
improvements.
Report pages. Report pages include visuals and filters to provide insights into your
data.
Dashboard and tiles. Dashboards and tiles offer an overview of the insights
included.
Sample data. A sample makes your app discoverable immediately after installation.

You might be familiar with each piece as existing Power BI features. When you build a
template app, there are other things to consider for each piece. For details, see the
following sections.

Queries
For template apps, queries developed in Power BI Desktop are used to connect to your
data source and import data. These queries are required to return a consistent schema
and are supported for Scheduled Data refresh.

Connect to your API

To get started, you need to connect to your API from Power BI Desktop to start building
your queries.

You can use the Data Connectors that are available in Power BI Desktop to connect to
your API. You can use the Web Data Connector (Get data > Web) to connect to your
Rest API or the OData connector (Get data > OData feed) to connect to your OData
feed.
 ７ Note

Currently, template apps do not support custom connectors. We recommend

exploring using Odatafeed Auth 2.0 as a mitigation for some of the connection
use-cases or to submit your connector for certification. For details on how to
develop a connector and certify it, see Data Connectors .

Consider the source

The queries define the data that's included in the data model. Depending on the size of
your system, these queries should also include filters to ensure your customers are
dealing with a manageable size that fits your business scenario.

Power BI template apps can run multiple queries in parallel and for multiple users
concurrently. Plan your throttling and concurrency strategy and ask us how to make
your template app fault tolerant.

Schema enforcement
Ensure your queries are resilient to changes in your system. Changes in schema on
refresh can break the model. If the source could return null or a missing schema result
for some queries, consider returning an empty table or a meaningful custom error
message.

Parameters
Parameters in Power BI Desktop allow your users to provide input values that
customize the data retrieved by the user. Think of the parameters up front to avoid
rework after investing time to build detailed queries or reports.

７ Note

Template apps support all parameters except Any and Binary .

Additional query tips

Ensure that all columns are typed appropriately.
Assign columns informative names. For more information, see Q&A.
For shared logic, consider using functions or queries.
 Privacy levels are currently not supported in the service. If you get a prompt about
privacy levels, you might need to rewrite the query to use relative paths.

Data models
A well-defined data model ensures that your customers can easily and intuitively
interact with the template app. Create the data model in Power BI Desktop.

７ Note

You should do much of the basic modeling, such as typing and column names, in
the queries.

Q&A
The modeling also affects how well Q&A can provide results for your customers. Be sure
to add synonyms to commonly used columns, and properly name your columns in the
queries.

Additional data model tips

Make sure you've:

Applied formatting to all value columns. Apply types in the query.

Applied formatting to all measures.
Set default summarization. In particular, set Do Not Summarize when applicable,
for unique values, for example.
Set a data category, when applicable.
Set relationships, as necessary.

Reports
The report pages offer extra insight into the data included in your template app. Use the
pages of the reports to answer the key business questions your template app is trying to
address. Create the report using Power BI Desktop.

Additional report tips

Use more than one visual per page for cross-filtering.
 Align the visuals carefully, with no overlapping.
Ensure that the page is set to 4:3 or 16:9 mode for layout.
Ensure that all of the aggregations presented make numeric sense, for instance,
averages or unique values.
Check that slicing produces rational results.
Include your logo on at least the top report.
Ensure that elements are in the client's color scheme to the extent possible.

７ Note

A single template app cannot include more than twenty reports.

Dashboards
The dashboard is the main point of interaction with your template app for your
customers. It should include an overview of the content included, especially the
important metrics for your business scenario.

To create a dashboard for your template app, just upload your PBIX through Get data >
Files or publish directly from Power BI Desktop.

Additional dashboard tips

Maintain the same theme when pinning so that the tiles on your dashboard are
consistent.
Pin a logo to the theme so consumers know where the pack is from.
Suggested layout to work with most screen resolutions is five to six small tiles
wide.
All dashboard tiles should have appropriate titles and subtitles.
Consider groupings in the dashboard for different scenarios, either vertically or
horizontally.

Sample data
A template app, as part of the app creation stage, wraps the cache data in the
workspace as part of the app:

Allows the installer to understand the functionality and purpose of the app before
connecting data.
 Creates an experience that drives the installer to further explore app capabilities,
which leads to connecting the app dataset.

We recommend having quality sample data before creating the app to ensure that the
app's report and dashboards are populated with data. Try to keep sample data size as
small as possible.

Publishing on AppSource
Template apps can be published on AppSource. Follow these guidelines before
submitting your app to AppSource:

Make sure that you create a template app with engaging sample data that can
help the installer understand what the app can do. Empty reports and dashboards
won't be approved.
Template apps support sample data only apps. Make sure to check the static app
checkbox.
Have instructions for the validation team to follow that include credentials and
parameters they can use to connect to the data.
Your application must include an App icon in Power BI and on your cloud partner
portal (CPP) offer.
Configure the landing page.
Make sure to follow the documentation about the Power BI App offer.
If a dashboard is part of your app, make sure it's not empty.
Install the app using the app link before submitting it. Make sure that you can
connect the dataset and that the app experience is as you planned.
Before uploading a PBIX file into the template workspace, make sure to unload any
unnecessary connections.
Follow Power BI Best design practices for reports and visuals to achieve maximum
impact on your users and getting approved for distribution.

Create a download link for the app

After publishing the template app on AppSource, consider creating a download link
from your website to either:

The AppSource download page, which can be viewed by publicly. Get the link from
your AppSource page.
Power BI, which can be viewed by a Power BI user.
In order to redirect a user to the app's download link in Power BI, see the following code
example: GitHub repo .

Automate parameter configuration during

installation
If you're an independent software vendor (ISV) and are distributing your template app
via your web service, you can create automation that configures template app
parameters automatically when your customers install the app in their Power BI account.
This approach makes things easier for your customers. It also increases the likelihood of
a successful installation because they don't have to supply details that they might not
know. For more information, see Automated configuration of a template app
installation.

Next steps
What are Power BI template apps?
Manage your published template app
Article • 11/10/2023

If you have a Power BI template app in production, when you want to make changes to
the app, you can start over in the test phase, without interfering with the app in
production.

Update your app

Go to the template app workspace. Then, if you made your changes in Power BI
Desktop, start at Step 1. If you did not make any changes in Power BI Desktop, start at
Step 2.

1. Upload your updated semantic model and make sure to overwrite the existing
semantic model.

If the .pbix file you're uploading has the same name as the semantic model
and report used in the app, uploading will overwrite the existing semantic
model.

If you're changing the name of the semantic model and report used in the
app, and the .pbix file you want to upload has a different name than the
semantic model and report used in the app, do the following:
Rename the semantic model and report used in the app so that their
names exactly match the name of your updated .pbix file.
Upload your .pbix file and overwrite the existing semantic model and
report that you just renamed.

In either case, to upload a local .pbix file to the service, select Upload >
Browse, navigate to the file, and select Open. A dialog will ask for your
permission to overwrite the semantic model the app uses. If you don't
overwrite the existing semantic model, customers won't be able to install
your updated app.
 ） Important

Never delete the semantic model the app uses. Deleting the semantic model
makes it impossible for customers to update their copies of the app.

2. In the Release management pane for the app, select Create app.

3. Repeat the app creation process. If you changed the name of the semantic model
and report used in the app, you might want to rename the app as well.

4. After you set Branding, Navigation, Control, Parameters, Authentication, and

Access, select Create app again to save your changes, and then select Close.

5. Select Release management again.

In the Release management pane, you now see two versions of the app: The
version in Production, plus a new version in Testing.
6. When you're ready to promote your app to pre-production for further testing
outside your tenant, go back to the Release Management pane and select
Promote app next to Testing.

You now have a version in Production and a version in Pre-production.

Your link is now live.

７ Note
 The Promote app button at the pre-production stage is disabled. Disabling
the button prevents accidentally overwriting the live production link with the
current app version before the Cloud Partner Portal (CPP) validates and
approves the new app version.

7. Submit your link again to the CPP by following the steps at Power BI App offer
update. In the CPP, you must publish your offer again and have it validated and
approved. If you've changed the name of the app, be sure to also change the
name in the CPP. When your offer is approved, the Promote app button becomes
active again.

8. Promote your app to the Production stage.

Update behavior
Updating the app lets template app installers update their template app in the
already installed workspaces without losing the connection configuration.

To learn how changes in the semantic model affect the installed template app, see
Overwrite behavior.

When a template app is overwritten and updates, it first reverts back to sample
data, and automatically reconnects using the installer's configuration parameters
and authentication. Until refresh is complete, the reports, dashboards, and
organizational app display the sample data banner.

If you added a new query parameter to the updated semantic model that requires
user input, you must select the Required checkbox. This selection prompts the
installer with the connection string after updating the app.
Extract workspace
It's easy to roll back to the previous version of a template app with the extract capability.
The following steps extract a specific app version from a release stage into a new
workspace:

1. In the Release Management pane, next to an app version, select More options (...)
and then select Extract.
2. In the confirmation dialog box, enter a name for the extracted workspace, and
select Extract. Power BI adds a new workspace for the extracted app.
Your new workspace versioning resets, and you can continue to develop and distribute
the template app from the newly extracted workspace.

Delete template app version

A template app workspace is the source of an active distributed template app. To
protect the template app users, it's not possible to delete a template app workspace
without first removing all the created app versions in the workspace. Deleting an app
version also deletes the app URL, which no longer works.

1. In the Release Management pane, next to the app version you want to delete,
select More options (...) and then select Delete.
2. In the confirmation dialog box, select Delete.
 ７ Note

Make sure not to delete app versions that customers or AppSource are using, or
they will no longer work.

Next steps
See how your customers interact with your template app in Install, customize, and
distribute template apps in your organization.
See the Power BI Application offer for details on distributing your app.
Install, share, and update template apps
in your organization
Article • 02/08/2023

Are you a Power BI analyst? Here you can learn more about template apps and how to
connect to many of the services that you use to run your business, such as Salesforce,
Microsoft Dynamics, and Google Analytics. You can then modify the template app's pre-
built dashboard and reports to suit the needs of your organization, and distribute them
to your colleagues as apps.

If you're interested in creating template apps yourself for distribution outside your
organization, see Create a template app in Power BI. With little or no coding, Power BI
partners can build Power BI apps and make them available to Power BI customers.

Prerequisites
To install, customize, and distribute a template app, you need:

A Power BI pro license.

Permissions to install template apps on your tenant.
A valid installation link for the app, which you get either from AppSource or from
the app creator.
A good familiarity with the basic concepts of Power BI.

Install a template app

1. In the nav pane in the Power BI service, select Apps > Get apps.

2. In the Power BI apps marketplace that appears, select Template apps. All the
template apps available in AppSource are shown. Browse to find the template app
you're looking for, or get a filtered selection by using the search box. Type part of
the name of the template app, or select a category such as finance, analytics, or
marketing to find the item you're looking for.

3. When you find the template app you're looking for, select it. The template app
offer appears. Select Get It Now.
 

4. In the dialog box that appears, select Install.

The app is installed, along with a workspace of the same name that has all the
artifacts needed for further customization.

７ Note

If you use an installation link for an app that isn't listed on AppSource, a
validation dialog box will ask you to confirm your choice.

To be able to install a template app that isn't listed on AppSource, you can
request the relevant permissions from your admin. See the template app
settings in Power BI admin portal for details.

When the installation finishes successfully, a notification tells you that your new
app is ready.
Connect to data
1. Select Go to app.

The app opens, showing sample data.

2. Select the Connect your data link on the banner at the top of the page.

This link opens the parameters dialog, where you change the data source from the
sample data to your own data source (see known limitations), followed by the
authentication method dialog. You might have to redefine the values in these
dialogs. See the documentation of the specific template app you're installing for
details.
 

Once you've finished filling out the connection dialogs, the connection process
starts. A banner informs you that the data is being refreshed, and that in the
meantime you're viewing sample data.

Your report data will automatically refresh once a day, unless you disabled this
setting during the sign-in process. You can also set up your own refresh schedule
to keep the report data up to date if you so desire.

Customize and share the app

After you've connected to your data and data refresh is complete, you can customize
any of the reports and dashboards the app includes, as well as share the app with your
colleagues. Remember, however that any changes you make will be overwritten when
you update the app with a new version, unless you save the items you changed under
different names. See details about overwriting.
To customize and share your app, select the pencil icon at the top right corner of the
page.

For information about editing artifacts in the workspace, see

Tour the report editor in Power BI

Basic concepts for designers in the Power BI service

When you're done making changes to the artifacts in the workspace, you're ready to
publish and share the app. See Publish your app to learn how.

Update a template app

From time to time, template app creators release new improved versions of their
template apps, via AppSource, a direct link, or both.

If you originally downloaded the app from AppSource, when a new version of the
template app becomes available, you get notified in two ways:

An update banner appears in the Power BI service informing you that a new app
version is available.

You receive a notification in Power BI's notification pane.

 ７ Note

If you originally got the app via a direct link rather than through AppSource, the
only way to know when a new version is available is to contact the template app
creator.

To install the update, either select Get it on the notification banner or in the notification
center, or find the app again in AppSource and choose Get it now. If you got a direct
link for the update from the Template app creator, select the link.

You're asked how you want the update to affect your currently installed app.
 Update the workspace and the app: Updates both the workspace and the app,
and republishes the app to your organization. Choose this option if you didn't
make any changes to the app or its content and want to overwrite the old app.
Your connections will be re-established, and the new version of the app will include
any updated app branding, such as app name, logo, and navigation, as well as the
latest publisher improvements to content.

Update only workspace content without updating the app: Updates the reports,
dashboards, and dataset in the workspace. After updating the workspace, you can
choose what you want to include in the app, and then you need to update the app
to republish it to your organization with the changes.

Install another copy of the app into a new workspace: Installs a fresh version of
the workspace and app. Choose this option if you don’t want to change your
current app.

Overwrite behavior
Overwriting updates the reports, dashboards, and dataset in the workspace, not
the app. Overwriting doesn't change app navigation, setup, and permissions.

If you chose the second option, after you've updated the workspace you need to
update the app to apply changes from the workspace to the app.

Overwriting keeps configured parameters and authentication. After the update, an

automatic dataset refresh starts. During this refresh, the app, reports, and
dashboards present sample data.

Overwriting always presents sample data until the refresh is complete. If the
template app author made changes to the dataset or parameters, users of the
workspace and app won't see the new data until the refresh is complete. Instead,
they'll still see sample data during this time.

Overwriting never deletes new reports or dashboards you've added to the

workspace. It only overwrites the original reports and dashboards with changes
from the original author.

） Important
 Remember to update the app after overwriting to apply changes to the reports and
dashboard for your organizational app users.

Delete a template app

An installed template app consists of the app and its associated workspace. If you want
to remove the template app, you have two options:

Completely remove the app and its associated workspace: To completely remove
a template app and its associated workspace, go to the app tile on the Apps page,
select the trash icon, and then choose Delete in the dialog that appears.

Unpublish the app: This option removes the app but keeps its associated
workspace. This option is useful if there are customizations that you made and
want to keep.

To unpublish the app:

1. Open the app.

2. Select the edit app pencil icon to open the template app's workspace.

3. In the template app workspace, select More options (...), and then choose
Unpublish App.

Next steps
Create a workspace in Power BI
Automated configuration of a template
app installation
Article • 11/29/2021

Template apps are a great way for customers to start getting insights from their data.
Template apps get them up and running quickly by connecting them to their data. The
template apps provide them with prebuilt reports that they can customize if they so
desire.

Customers aren't always familiar with the details of how to connect to their data. Having
to provide these details when they install a template app can be a pain point for them.

If you are a data services provider and have created a template app to help your
customers get started with their data on your service, you can make it easier for them to
install your template app. You can automate the configuration of your template app's
parameters. When the customer signs in to your portal, they select a special link you've
prepared. This link:

Launches the automation, which gathers the information it needs.

Preconfigures the template app parameters.
Redirects the customer to their Power BI account where they can install the app.

All they have to do is select Install and authenticate against their data source, and
they're good to go!

The customer experience is illustrated here.

This article describes the basic flow, the prerequisites, and the main steps and APIs you
need to automate the configuration of a template app installation. If you want to dive in
and get started, you can skip to the tutorial where you automate the configuration of
the template app installation by using a simple sample application we've prepared that
uses an Azure function.

Basic flow
The basic flow of automating the configuration of a template app installation is as
follows:

1. The user signs in to the ISV's portal and selects the supplied link. This action
initiates the automated flow. The ISV's portal prepares the user-specific
configuration at this stage.

2. The ISV acquires an app-only token based on a service principal (app-only token)
that's registered in the ISV's tenant.

3. Using Power BI REST APIs, the ISV creates an install ticket, which contains the user-
specific parameter configuration as prepared by the ISV.

4. The ISV redirects the user to Power BI by using a POST redirection method that
contains the install ticket.

5. The user is redirected to their Power BI account with the install ticket and is
prompted to install the template app. When the user selects Install, the template
app is installed for them.

７ Note

While parameter values are configured by the ISV in the process of creating the
install ticket, data source-related credentials are only supplied by the user in the
final stages of the installation. This arrangement prevents them from being exposed
to a third party and ensures a secure connection between the user and the
template app data sources.

Prerequisites
To provide a preconfigured installation experience for your template app, the following
prerequisites are required:
 A Power BI Pro license. If you're not signed up for Power BI Pro, sign up for a free
trial before you begin.

Your own Azure Active Directory (Azure AD) tenant set up. For instructions on how
to set one up, see Create an Azure AD tenant.

A service principal (app-only token) registered in the preceding tenant. For more
information, see Embed Power BI content with service principal and an application
secret. Make sure to register the application as a server-side web application app.
You register a server-side web application to create an application secret. From this
process, you need to save the application ID (ClientID) and application secret
(ClientSecret) for later steps.

A parameterized template app that's ready for installation. The template app must
be created in the same tenant in which you register your application in Azure AD.
For more information, see Template app tips or Create a template app in Power BI.
From the template app, you need to note the following information for the next
steps:
App ID, Package Key, and Owner ID as they appear in the installation URL at the
end of the process of defining the properties of the template app when the app
was created. You can also get the same link by selecting Get link in the template
app's Release Management pane.
Parameter names as they're defined in the template app's dataset. Parameter
names are case-sensitive strings and can also be retrieved from the Parameter
Settings tab when you define the properties of the template app or from the
dataset settings in Power BI.

To be able to test your automation work flow, add the service principal to the
template app workspace as an Admin.

７ Note

You can test your preconfigured installation application on your template app
if the template app is ready for installation, even if it isn't publicly available on
AppSource yet. For users outside your tenant to be able to use the automated
installation application to install your template app, the template app must be
publicly available in the Power BI apps marketplace . Before you distribute
your template app by using the automated installation application you're
creating, be sure to publish it to Partner Center.

Main steps and APIs

The main steps for automating the configuration of a template app installation, and the
APIs you'll need, are described in the following sections. While most of the steps are
done with Power BI REST APIs, the code examples described here are made with the
.NET SDK.

Step 1: Create a Power BI client object

Using Power BI REST APIs requires you to get an access token for your service principal
from Azure AD. You're required to get an Azure AD access token for your Power BI
application before you make calls to the Power BI REST APIs. To create the Power BI
client with your access token, you need to create your Power BI client object, which
allows you to interact with the Power BI REST APIs. You create the Power BI client object
by wrapping the AccessToken with a Microsoft.Rest.TokenCredentials object.

C#

using Microsoft.IdentityModel.Clients.ActiveDirectory;
using Microsoft.Rest;
using Microsoft.PowerBI.Api.V2;

var tokenCredentials = new

TokenCredentials(authenticationResult.AccessToken, "Bearer");

// Create a Power BI client object. It's used to call Power BI APIs.

using (var client = new PowerBIClient(new Uri(ApiUrl), tokenCredentials))
{
// Your code goes here.
}

Step 2: Create an install ticket

Create an install ticket, which is used when you redirect your users to Power BI. The API
used for this operation is the CreateInstallTicket API.

Template Apps CreateInstallTicket

A sample of how to create an install ticket for template app installation and
configuration is available from the InstallTemplateApp/InstallAppFunction.cs file in the
sample application .

The following code example shows how to use the template app CreateInstallTicket
REST API.

C#
 using Microsoft.PowerBI.Api.V2;
using Microsoft.PowerBI.Api.V2.Models;

// Create Install Ticket Request.

InstallTicket ticketResponse = null;
var request = new CreateInstallTicketRequest()
{
InstallDetails = new List<TemplateAppInstallDetails>()
{
new TemplateAppInstallDetails()
{
AppId = Guid.Parse(AppId),
PackageKey = PackageKey,
OwnerTenantId = Guid.Parse(OwnerId),
Config = new TemplateAppConfigurationRequest()
{
Configuration = Parameters
.GroupBy(p => p.Name)
.ToDictionary(k => k.Key, k =>
k.Select(p => p.Value).Single())
}
}
}
};

// Issue the request to the REST API using .NET SDK.

InstallTicket ticketResponse = await
client.TemplateApps.CreateInstallTicketAsync(request);

Step 3: Redirect users to Power BI with the

ticket
After you've created an install ticket, you use it to redirect your users to Power BI to
continue with the template app installation and configuration. You use a POST method
redirection to the template app's installation URL, with the install ticket in its request
body.

There are various documented methods of how to issue a redirection by using POST
requests. Choosing one or another depends on the scenario and how your users interact
with your portal or service.

A simple example, mostly used for testing purposes, uses a form with a hidden field,
which automatically submits itself upon loading.

JavaScript

<html>
<body onload='document.forms["form"].submit()'>
 <!-- form method is POST and action is the app install URL -->
<form name='form' action='https://app.powerbi.com/....'
method='post' enctype='application/json'>
<!-- value should be the new install ticket -->
<input type='hidden' name='ticket' value='H4sI....AAA='>
</form>
</body>
</html>

The following example of the sample application 's response holds the install ticket
and automatically redirects users to Power BI. The response for this Azure function is the
same automatically self-submitting form that we see in the preceding HTML example.

C#

...
return new ContentResult() { Content = RedirectWithData(redirectUrl,
ticket.Ticket), ContentType = "text/html" };
}

...

public static string RedirectWithData(string url, string ticket)

{
StringBuilder s = new StringBuilder();
s.Append("<html>");
s.AppendFormat("<body onload='document.forms[\"form\"].submit()'>");
s.AppendFormat("<form name='form' action='{0}' method='post'
enctype='application/json'>", url);
s.AppendFormat("<input type='hidden' name='ticket' value='{0}' />",
ticket);
s.Append("</form></body></html>");
return s.ToString();
}

７ Note

There are various methods of using POST browser redirects. You should always use
the most secure method, which depends on your service needs and restrictions.
Remember that some forms of insecure redirection can result in exposing your
users or service to security issues.

Step 4: Move your automation to production

When the automation you've designed is ready, be sure to move it to production.
Next steps
Try our tutorial, which uses a simple Azure function to automate the configuration
of a template app installation.
More questions? Try asking the Power BI Community .
Connect to the services you use with
Power BI
Article • 01/09/2023

With Power BI, you can connect to many of the services you use to run your business,
such as Salesforce, Microsoft Dynamics, and Google Analytics. Power BI starts by using
your credentials to connect to the service. It creates a Power BI workspace with a
dashboard and a set of Power BI reports that automatically show your data and provide
visual insights about your business.

Sign in to Power BI to view all of the services you can connect to. Select Apps > Get
apps.

After you install the app, you can view the dashboard and reports in the app and the
workspace in the Power BI service (https://powerbi.com ). You can also view them in
the Power BI mobile apps. In the workspace, you can modify the dashboard and reports
to meet the needs of your organization, and then distribute them to your colleagues as
an app.
Get started
1. Select Apps in the navigation pane, then choose Get apps in the upper-right
corner.

2. In Power BI apps, select the Apps tab, and search for the service you want.
Edit the dashboard and reports
When the import is complete, the new app appears on the Apps page.

1. Select Apps in the navigation pane, then choose the app.

2. You can ask a question by typing in the Q&A box, or select a tile to open the
underlying report.
 Change the dashboard and report to fit the needs of your organization. Then
distribute your app to your colleagues

What's included
After connecting to a service, you see a newly created app and workspace with a
dashboard, reports, and dataset. The data from the service is focused on a specific
scenario and might not include all the information from the service. The data is
scheduled to refresh automatically once per day. You can control the schedule by
selecting the dataset.

You can also connect to many services in Power BI Desktop, such as Google Analytics,
and create your own customized dashboards and reports.

For more information on connecting to specific services, see the individual help pages.

Troubleshooting
Empty tiles: While Power BI is first connecting to the service, you might see an empty
set of tiles on your dashboard. If you still see an empty dashboard after two hours, it's
likely the connection failed. If you didn't see an error message with information on
correcting the issue, file a support ticket.

Select the question mark icon (?) in the upper-right corner > Get help.
Missing information: The dashboard and reports include content from the service
focused on a specific scenario. If you're looking for a specific metric in the app and don't
see it, add an idea on the Power BI Support page.

Suggesting services
Do you use a service you'd like to suggest for a Power BI app? Go to the Power BI
Support page and let us know.

If you're interested in creating template apps to distribute yourself, see Create a

template app in Power BI. Power BI partners can build Power BI apps with little or no
coding, and deploy them to Power BI customers.

Next steps
Distribute apps to your colleagues
Create workspaces in Power BI
Questions? Try asking the Power BI Community
Connect to the COVID-19 US tracking
report
Article • 12/09/2020

This article tells you how to install the template app for the COVID-19 tracking report,
and how to connect to the data sources.

For detailed information about the report itself, including disclaimers and information
about the data, see COVID-19 tracking sample for US state and local governments.

After you've installed the template app and connected to the data sources, you can
customize the report as per your needs. You can then distribute it as an app to
colleagues in your organization.

Install the app

1. Click the following link to get to the app: COVID-19 US Tracking Report template
app

2. Once you're on the App's AppSource page, click GET IT NOW .

 3. When prompted , click Install. Once the app has installed, you will see it on your
Apps page.

Connect to data sources

1. Click the icon on your Apps page to open the app. The app opens, showing sample
data.

2. Select the Connect your data link on the banner at the top of the page.
3. The parameters dialog will appear. There are no required parameters. Click Next.

4. The authentication method dialog will appear. Recommended values are

prepopulated. Don't change these unless you have specific knowledge of different
values.

Click Next.
5. Click Sign in.
 The report will connect to the data sources and be populated with up-to-date
data. During this time you will see sample data and that refresh is in progress.

Schedule report refresh

When the data refresh has completed, you will be in the workspace associated with the
app. Set up a refresh schedule to keep the report data up to date.
Customize and share
See Customize and share the app for details. Be sure to review the report disclaimers
before publishing or distributing the app.

Next steps
COVID-19 tracking sample for US state and local governments
Questions? Try asking the Power BI Community
What are Power BI template apps?
Install and distribute template apps in your organization
Connect to the Crisis Communication
Presence Report
Article • 12/09/2020

This Power BI app is the report/dashboard artifact in the Microsoft Power Platform
solution for Crisis Communication. It tracks worker location for Crisis Communication
app users. The solution combines capabilities of Power Apps, Power Automate, Teams,
SharePoint and Power BI. It can be used on the web, mobile or in Teams.

The dashboard shows emergency managers aggregate data across their health system
to help them to make timely, correct decisions.

This article tells you how to install the app and how to connect to the data sources. For
more information about the Crisis Communication app, see Set up and learn about the
Crisis Communication sample template in Power Apps

After you've installed the template app and connected to the data sources, you can
customize the report as per your needs. You can then distribute it as an app to
colleagues in your organization.

Prerequisites
Before installing this template app, you must first install and set up the Crisis
Communication sample. Installing this solution creates the datasource references
necessary to populate the app with data.
When installing the Crisis Communication sample, take note of the SharePoint list folder
path of "CI_Employee Status" and list ID.

Install the app

1. Click the following link to get to the app: Crisis Communication Presence Report
template app

2. On the AppSource page for the app, select GET IT NOW .

3. Read the information in One more thing, and select Continue.

4. Select Install.
 Once the app has installed, you see it on your Apps page.

Connect to data sources

1. Select the icon on your Apps page to open the app.

The app opens, showing sample data.

2. Select the Connect your data link on the banner at the top of the page.
3. In the dialog box:
a. In the SharePoint_Folder field, enter your "CI_Employee Status" SharePoint list
path.
b. In the List_ID field, enter your list ID that you got from list settings. When done,
click Next.

4. In the next dialog that appears, set the authentication method to OAuth2. You
don't have to do anything to the privacy level setting.

Select Sign in.

5. At the Microsoft sign-in screen, sign in to Power BI.

After you've signed in, the report connects to the data sources and is populated
with up-to-date data. During this time, the activity monitor turns.
Schedule report refresh
When the data refresh has completed, set up a refresh schedule to keep the report data
up to date.

1. In the top header bar, select Power BI.

2. In the left navigation pane, look for the Hospital Emergency Response Decision
Support Dashboard workspace under Workspaces, and follow the instruction
described in the Configure scheduled refresh article.

Customize and share

See Customize and share the app for details. Be sure to review the report disclaimers
before publishing or distributing the app.

Next steps
Set up and learn about the Crisis Communication sample template in Power Apps
Questions? Try asking the Power BI Community
What are Power BI template apps?
Install and distribute template apps in your organization
Connect to GitHub with Power BI
Article • 09/19/2022

This article walks you through pulling your data from your GitHub account with a Power
BI template app. The template app generates a workspace with a dashboard, a set of
reports, and a dataset to allow you to explore your GitHub data. The GitHub app for
Power BI shows you insights into your GitHub repository, also known as repo, with data
around contributions, issues, pull requests, and active users.

After you've installed the template app, you can change the dashboard and report. Then
you can distribute it as an app to colleagues in your organization.

Connect to the GitHub template app or read more about the GitHub integration
with Power BI.

You can also try the GitHub tutorial. It installs real GitHub data about the public repo for
the Power BI documentation.

７ Note

This template app requires the GitHub account to have access to the repo. More
details on requirements below.

This template app does not support GitHub Enterprise.

Install the app

1. Click the following link to get to the app: GitHub template app

2. On the AppSource page for the app, select GET IT NOW .

3. Select Install.

Once the app has installed, you see it on your Apps page.
Connect to data sources
1. Select the icon on your Apps page to open the app.

The app opens, showing sample data.

2. Select the Connect your data link on the banner at the top of the page.

3. This opens the parameters dialog, where you change the data source from the
sample data to your own data source (see known limitations), followed by the
authentication method dialog. You may have to redefine the values in these
dialogs.
4. Enter your GitHub credentials and follow the GitHub authentication process (this
step might be skipped if you're already signed in with your browser).
Once you've finished filling out the connection dialogs and signed in to GitHub, the
connection process starts. A banner informs you that the data is being refreshed, and
that in the meantime you are viewing sample data.

Your report data will automatically refresh once a day, unless you disabled this during
the sign-in process. You can also set up your own refresh schedule to keep the report
data up to date if you so desire.

Customize and share

To customize and share your app, select the pencil icon at the top right corner of the
page.

For information about editing items in the workspace, see

Tour the report editor in Power BI

Basic concepts for designers in the Power BI service
Once you are done making any changes you wish to the items in the workspace, you are
ready to publish and share the app. See Create and publish your app to learn how to do
this.

What's included in the app

The following data is available from GitHub in Power BI:

Table name Description

Contributions The contributions table gives the total additions, deletions, and
commits authored by the contributor aggregated per week. The top
100 contributors are included.

Issues List all issues for the selected repo and it contains calculations like total
and average time to close an issue, Total open issues, Total closed
issues. This table will be empty when there are no issues in the repo.

Pull requests This table contains all the Pull Requests for the repo and who pulled the
request. It also contains calculations around how many open, closed,
and total pull requests, how long it took to pull the requests and how
long the average pull request took. This table will be empty when there
are no pull requests in the repo.

Users This table provides a list of GitHub users or contributors who have
made contributions, filed issues, or solved Pull requests for the repo
selected.

Milestones It has all the Milestones for the selected repo.

DateTable This table contains dates from today and for years in the past that allow
you to analyze your GitHub data by date.

ContributionPunchCard This table can be used as a contribution punch card for the selected
repo. It shows commits by day of week and hour of day. This table is
not connected to other tables in the model.

RepoDetails This table provides details for the repo selected.

System requirements
The GitHub account that has access to the repo.
Permission granted to the Power BI for GitHub app during first login. See details
below on revoking access.
Sufficient API calls available to pull and refresh the data.
 ７ Note

This template app does not support GitHub Enterprise.

De-authorize Power BI
To de-authorize Power BI from being connected to your GitHub repo, you can Revoke
access in GitHub. See this GitHub help topic for details.

Finding parameters
You can determine the owner and repository by looking at the repository in GitHub
itself:

The first part "Azure" is the owner and the second part "azure-sdk-for-php" is the
repository itself. You see these same two items in the URL of the repository:

Console

<https://github.com/Azure/azure-sdk-for-php> .

Troubleshooting
If necessary, you can verify your GitHub credentials.

1. In another browser window, go to the GitHub web site and sign in to GitHub. You
can see you’re logged in, in the upper-right corner of the GitHub site.
2. In GitHub, navigate to the URL of the repo you plan to access in Power BI. For
example: https://github.com/dotnet/corefx .
3. Back in Power BI, try connecting to GitHub. In the Configure GitHub dialog box,
use the names of the repo and repo owner for that same repo.
Next steps
Tutorial: Connect to a GitHub repo with Power BI
Create workspaces in Power BI
Install and use apps in Power BI
Connect to Power BI apps for external services
Questions? Try asking the Power BI Community
Connect to the Hospital Emergency
Response Decision Support Dashboard
Article • 11/30/2020

The Hospital Emergency Response Decision Support Dashboard template app is the
reporting component of the Microsoft Power Platform solution for healthcare
emergency response . The dashboard shows emergency managers aggregate data
across their health system to help them to make timely, correct decisions.

This article tells you how to install the app and how to connect to the data sources. To
learn how to use the report that you will see with this app, see the Hospital Emergency
Response Decision Support Dashboard documentation.

After you've installed the template app and connected to the data sources, you can
customize the report as per your needs. You can then distribute it as an app to
colleagues in your organization.

Prerequisites
Before installing this template app, you must first install and set up the Hospital
Emergency Response Power Platform solution. Installing this solution creates the
datasource references necessary to populate the app with data.

When installing Hospital Emergency Response Power Platform solution, take note of the
URL of your Common Data Service environment instance. You will need it to connect the
template app to the data.
Install the app
1. Click the following link to get to the app: Hospital Emergency Response Decision
Support Dashboard template app

2. On the AppSource page for the app, select GET IT NOW .

3. Read the information in One more thing, and select Continue.

4. Select Install.
 Once the app has installed, you see it on your Apps page.

Connect to data sources

1. Select the icon on your Apps page to open the app.

2. On the splash screen, select Explore.

 The app opens, showing sample data.

3. Select the Connect your data link on the banner at the top of the page.

4. In the dialog box:

a. In the organization name field, enter the name of your organization, for
example, "Contoso Health Systems". This field is optional. This name appears in
the upper-left side of the dashboard.
b. In the CDS_base_solution field, Type the URL of your Common Data Service
environment instance. For example: https://[myenv].crm.dynamics.com. When
done, click Next.
5. In the next dialog that appears, set the authentication method to OAuth2. You
don't have to do anything to the privacy level setting.

Select Sign in.

6. At the Microsoft sign-in screen, sign in to Power BI.

 After you've signed in, the report connects to the data sources and is populated
with up-to-date data. During this time, the activity monitor turns.

Schedule report refresh

When the data refresh has completed, set up a refresh schedule to keep the report data
up to date.

1. In the top header bar, select Power BI.

 2. In the left navigation pane, look for the Hospital Emergency Response Decision
Support Dashboard workspace under Workspaces, and follow the instructions
described in the Configure scheduled refresh article.

Customize and share

See Customize and share the app for details. Be sure to review the report disclaimers
before publishing or distributing the app.

Next steps
Understanding the Hospital Emergency Response report
Set up and learn about the Crisis Communication sample template in Power Apps
Questions? Try asking the Power BI Community
What are Power BI template apps?
Install and distribute template apps in your organization
Connect to the Emissions Impact
Dashboard for Azure
Article • 03/14/2024

Calculate your cloud-based carbon emissions today with the Emissions Impact
Dashboard for Azure.

Accurate carbon accounting requires good information from partners, vendors, and
suppliers. The Emissions Impact Dashboard for Azure gives you transparency on the
carbon emissions generated by your usage of Azure and Microsoft Dynamics.
Microsoft’s carbon accounting extends across all three scopes of emissions with a
methodology validated by Stanford University in 2018. It uses consistent and accurate
carbon accounting to quantify the impact of Microsoft cloud services on customers’
environmental footprint. Microsoft is the only cloud provider to provide this level of
transparency to customers while compiling reports for voluntary or statutory reporting
requirements.

） Important

In February 2024, your data will undergo recalculation due to an update in the
methodology that now allows for a more detailed attribution of carbon emissions.
For more information about these changes, go to Calculations update FAQ.

Prerequisites
To install the Emissions Impact Dashboard for Azure in Power BI and connect it to your
data, make sure you have a Power BI Pro license. If you don’t have a Power BI Pro
license, get a free trial now .

The Emissions Impact Dashboard for Azure is supported for EA Direct, MCA, and MPA
accounts with direct billing relationships with Microsoft.

If you have an EA Direct account, you must be a Billing Account Administrator

(formerly known as an Enrollment Administrator) with either read or write
permissions and have your company's billing account ID (formerly known as the
enrollment number).

If you have an MCA or MPA and direct billing relationship with Microsoft, you must
be a Billing Account Administrator with a role as Billing Account
 Reader/Contributor/Owner and have your company's billing account ID.

） Important

Cloud solution providers (CSPs) are supported. Customers who purchase Azure
from a CSP aren't supported and must work directly with their CSP partner to learn
about their cloud emissions. Legacy accounts and China enrollments aren't
supported.

Install the app

1. Select the following link to get to the app: Emissions Impact Dashboard template
app .

2. On the AppSource page for the app, select GET IT NOW.

You can also search for the app in Power BI.

3. When prompted, select Install.

4. When the app finishes installing, it appears on your Power BI Apps page. Select the
app and open it.

5. Select Connect your data.

6. In the Connect to Emissions Impact Dashboard dialog that appears, under
EnrollmentIDorBillingAccountID, enter either your billing account ID (formerly
known as the enrollment number) for EA Direct customers or billing account ID for
MCA/MPA.

When done, select Next.

7. Connect your account:

For Authentication method, select OAuth2.

For Privacy level setting for this data source, select Organizational.
When done, select Sign in and connect.
8. Select the user account. Be sure to sign in with the credentials that have access to
the enrollmentID/Billing AccountID with valid permissions as explained in the
prerequisites.
 9. Wait for the view to build, which can take up to 24 hours. Refresh the dataset after
24 hours.

Update the app

Periodically you might receive update notifications from Appsource/Power BI about a
new version of the app. When you install the new version, the following options are
available:

Select Update the workspace and the app, and then select Install. The update installs,
overwriting the existing/installed workspace and app.

Issues
If there are any issues with the dataset refresh/app update during the updating process,
validate these steps and refresh the dataset.

Follow these steps to make sure your dataset configurations are set correctly:

1. Go to the workspace panel and open the app workspace.

2. Open the Scheduled Refresh option in the dataset settings and make sure the
billing account ID is correct.
3. Open the Parameters section and configure the data source again in the Data
Source section with the credentials that have access to the Enrollment ID / Billing
Account ID with valid permissions, mentioned in the prerequisites.

4. After the above steps are validated, go back to the app workspace and select the
Refresh option.
 5. After the dataset refreshes successfully, select the Update App option at the top-
right corner of the app workspace.

Additional resources
How-to video
The carbon benefits of cloud computing: A study on the Microsoft Cloud in
partnership with WSP

Finding your company's billing account ID

Follow these steps to find your company's billing account ID, or ask your organization’s
Azure administrator.

1. In the Azure portal , navigate to Cost Management + Billing.

2. In the Billing Scopes menu, select your billing account.

 3. Under Settings, select Properties. Your billing account ID displays under Billing
account.

Calculations update FAQ

Why are there updated values for my organization’s

Azure emissions data?
In January 2024, we refined our methodology for attributing carbon estimates. This
update allows for a more granular and precise allocation of carbon emissions to each
Azure resource, subscription, and customer.

What is the difference between the old and new

methodologies?
The new methodology now allows for a more enhanced and detailed attribution of
carbon emissions when a resource belongs to a nonspecific region such as All, Null, or
Global. Instead of only subscription-level carbon emissions data, you can now access
emissions information for each of your individual Azure resources. This granular data is
accessible through the Cloud for Sustainability export APIs, Project ESG lake, and Azure
Carbon Optimization capabilities. It enhances transparency and control over
environmental impact.
Is there a plan to update the Emissions Impact Dashboard
for Azure to display granular, resource-grain emissions
data?
At this time, we aren't enhancing the Emissions Impact Dashboard for Azure to display
resource-level emissions data. The dashboard maintains its current functionality,
continuing to present data at the subscription level. This data, powered by our updated
methodology, aggregates all resource-level data in the background and displays it at
the subscription level for a comprehensive view.

Has all the historical data been updated to reflect the

new methodology?
Yes, all historical Azure emissions data for your organization is recalculated using the
new methodology.

Is the data on the Emissions Impact Dashboard for Azure

and Cloud for Sustainability API the same?
The Emissions Impact Dashboard and the Cloud for Sustainability API provide data
based on the same source, and are the same.

Is there an option for me to access my Azure emissions

data using the old methodology?
Unfortunately, it isn't possible for us to provide the emissions data using the old
methodology.

General FAQs

App setup
I’m receiving an error at the time of connecting my data with the dashboard. What
can I do?

First, check Microsoft Cost Management and verify that you have Admin privileges. If
you don’t, request this access from your administrator. Next, ensure you’re using the
correct billing account ID or enrollment number.
I entered my enrollment number/billing account ID, but my company data isn’t
loading. What’s the issue?

The Emissions Impact Dashboard for Azure might take up to 24 hours to load your data.
Return after 24 hours and select the Refresh button in Power BI.

Is Microsoft trying to shift responsibility for emissions from Microsoft to me?

No. Carbon emissions from Azure services are reported as Microsoft's scope 1 and 2
emissions, consistent with the industry-standard Greenhouse Gas (GHG) Protocol . The
GHG Protocol defines scope 3 emissions as emissions another entity emits on your
behalf, and are inherently double-counted. The Emissions Impact Dashboard for Azure
provides new transparency to your scope 3 emissions associated with the use of Azure
services, specifically Scope 3 Category 1 "Purchased goods and services."

Why are my emissions from use of the Microsoft cloud so much lower than they
would be if I were using an on-premises solution?

Microsoft conducted a study, published in 2018 that evaluated the difference between
the Microsoft cloud and on-premises or traditional datacenters. The results show that
Azure Compute and Storage are between 52 and 79 percent more energy-efficient than
traditional enterprise datacenters. These numbers depend on the specific comparison
too low, medium, or high efficiency on-premises alternative being made. If you take into
account our renewable energy purchases, Azure is between 79 and 98 percent more
carbon efficient. These savings are due to four key features of the Microsoft Cloud: IT
operational efficiency, IT equipment efficiency, datacenter infrastructure efficiency, and
renewable electricity.

If Microsoft's operations are carbon neutral and powered by renewables, why aren't
customer emissions from Azure services zero?

There are two primary reasons why customer emissions from Microsoft aren’t zero. The
first is related to GHG accounting practices, and the second has to do with the boundary
of this analysis. To achieve carbon neutral operations, Microsoft uses carbon offsets to
reduce certain emission sources such as onsite fuel combustion for backup generators,
refrigerants, and vehicle fleets. These reduce Microsoft’s net emissions to zero. The
dashboard reports gross GHG emissions before the application of these offsets, though
the volume of offsets applied and net emissions is reported in the GHG Reporting tab
for further transparency. The second reason is that in addition to the energy and
emissions associated with the operation of Microsoft's datacenters, the emissions
footprint includes the energy used by Internet Service Providers outside of Microsoft’s
operational boundary to transmit data between Microsoft datacenters and Azure
customers.
How am I supposed to use this data, and where do I report it?

Your emissions can be reported as part of your company's Scope 3 indirect carbon
emissions. Scope 3 emissions are often disclosed in sustainability reports, CDP climate
change, and other reporting outlets. In addition to the emissions totals, the emissions
savings provide a clear example of how your company's decision to use Microsoft Azure
services is contributing to global emissions reductions. To contextualize, the app
indicates the equivalent vehicle miles avoided corresponding to the reduction in GHG
emissions, based on EPA’s equivalency calculator factors as of January 2020.

What can I do to reduce emissions further?

Being resource and cost efficient in Azure reduces the environmental impact from your
use of Azure. As an example, unused virtual machines are wasteful whether in the cloud
or on-premises. Right-sizing virtual machines to improve compute utilization factors
(CUF) decreases energy use per useful output, just as it does with physical servers.
Microsoft Cost Management and Azure carbon optimization give you the tools to plan
for, analyze and reduce your spending to maximize your cloud investment. The
sustainability guidance within the Azure Well-Architected Framework (WAF) is also
designed to help you optimize your cloud workloads and reduce your operational
footprint.

My company contract renewal process is underway and we'll have a new account
number. Will I lose my historical emissions data?

Yes, you will. Before your renewal, be sure to download all historical data and reports
you need for your records.

Can I export emissions data to Microsoft Excel?

You can export data from the GHG Preparation report, Usage report, and dashboard
page on a per-visualization level. You can't export the overall report’s data from the
Export button on the top header.

７ Note

Export to Excel is limited to 150,000 rows, and export to CSV is limited to 30,000
rows.

Methodology
What is the methodology behind the tool?
The Emissions Impact Dashboard for Azure reflects the specific cloud services consumed
and the associated energy requirements, efficiency of the datacenters providing those
services, electricity fuel mixes in the regions in which those datacenters operate, and
Microsoft’s purchases of renewable energy.

As part of the app’s development, the methodology and its implementation went
through third-party verification to ensure that it aligns to the World Resources Institute
(WRI)/World Business Council for Sustainable Development (WBCSD) Greenhouse Gas
(GHG) Protocol Corporate Accounting and Reporting Standard. The scope of the
verification, conducted in accordance with ISO 14064-3: Greenhouse gases--Part 3:
Specification with guidance for the validation and verification of greenhouse gas
assertions, included the estimation of emissions from Azure services, but excluded the
estimation of on-premises emissions given the counterfactual nature of that estimate. A
more detailed description of the carbon calculation is documented in the Calculation
Methodology tab in the tool.

What data is required to calculate the Azure carbon footprint? Do you access my
company's data?

The estimated carbon calculations are performed based on consumption of Azure

services accessed using Azure Consumed Revenue. The dashboard doesn't access any of
your stored customer data. The consumption data is combined with Microsoft's energy
and carbon tracking data to compute the estimated emissions associated with your
consumption of Azure services based on the datacenters that provide those services.

Does this calculation include all Azure services and all Azure regions?

The estimates include all Azure services in all Azure regions associated with the tenant
ID provided during setup.

Characterizing on-premises emissions

Where does the Emissions Impact Dashboard for Azure obtain data about my on-
premises emissions and operations?

The Emissions Impact Dashboard for Azure doesn’t obtain any information specifically
about your on-premises datacenters except what you provide. As described in
subsequent FAQs, the Emissions Impact Dashboard for Azure relies on industry research
and user inputs about the efficiency and energy mix of on-premises alternatives to
develop an estimate of on-premises emissions.

What are the assumptions regarding on-premises estimations? Are efficiency savings
just from improvements in Power Usage Effectiveness (PUE)?
Efficiencies associated with Microsoft cloud services include far more than improved
PUE. While Microsoft datacenters strive to optimize PUE, the primary efficiency
improvements come from IT operational efficiency (dynamic provisioning, multitenancy,
server utilization) and IT equipment efficiency (tailoring hardware to services ensuring
more energy goes towards useful output), in addition to datacenter infrastructure
efficiency (PUE improvements). Our 2018 study quantifies these savings compared to
a range of on-premises alternatives ranging from low-efficiency to high-efficiency
datacenters. These findings are used to estimate the energy use required for a
corresponding on-premises datacenter to provide the same services that each customer
consumes on the Microsoft cloud.

What is the assumed energy mix for the on-premises infrastructure?

By default, the Emissions Impact Dashboard for Azure estimates on-premises emissions
based on the mix of renewables and nonrenewables on the grid. It's assumed that the
on-premises datacenter would be located on the same grid as Microsoft’s datacenters.
However, for customers who purchase renewable electricity in addition to what’s on the
grid (for example, through Power Purchase Agreements), users can select the
percentage of renewable electricity, and the Emissions Impact Dashboard for Azure
adjusts on-premises emissions accordingly.

When should I choose Low, Medium, or High for the efficiency of the on-premises
infrastructure?

Users should select the efficiency most representative of the on-premises deployment
they would like to compare against, based on the equipment and datacenter
characteristics here:

Low: Physical servers and direct attached storage in small localized datacenter
(500-1,999 square feet)
Medium: Mix of physical/virtualized servers and attached/dedicated storage in
mid-tier internal datacenter (2,000-19,999 square feet)
High: Virtualized servers and dedicated storage in high-end internal datacenter
(>20,000 square feet)
Connect to the Emissions Impact
Dashboard for Microsoft 365
Article • 12/20/2023

Calculate emissions from your tenant's usage of Microsoft 365 with the Emissions
Impact Dashboard for Microsoft 365.

Accurate carbon accounting requires good information from partners, vendors, and
suppliers. The Emissions Impact Dashboard for Microsoft 365 gives you transparency on
the carbon emissions generated by your organization's usage of Microsoft 365.
Microsoft's carbon accounting extends across all three scopes of emissions with a third
party-validated methodology.

Prerequisites
To install the Emissions Impact Dashboard for Microsoft 365 in Power BI and connect it
to your data, make sure you have the following:

A business, enterprise, or education subscription for Microsoft 365 or Office 365.

A Power BI Pro license. If you don't have a Power BI Pro license, get a free trial
now .
One of the following Microsoft 365 admin roles:
Global admin
Exchange admin
Skype for Business admin
SharePoint admin
Global reader
Report reader

７ Note

In order to ensure that the report successfully refreshes over time, the Microsoft
365 admin credentials of the user who connects the application to your
organization's data must persist over time. If that user has the admin role removed
after the connection is established, the report will render demo data upon the next
refresh.

７ Note
 The Emissions Impact Dashboard for Microsoft 365 is currently not supported for
national/regional cloud deployments including but not limited to Microsoft's US
Government clouds and Office 365 operated by 21Vianet.

Included Microsoft 365 applications

The Emissions Impact Dashboard for Microsoft 365 reports datacenter emissions and
active usage associated with the following applications:

Exchange Online
SharePoint
OneDrive
Microsoft Teams
Word
Excel
PowerPoint
Outlook

Install the app

1. Select the following link to get to the app: Emissions Impact Dashboard for
Microsoft 365 .

2. On the AppSource page for the app, select GET IT NOW. You can also search for
the app in Power BI.

3. When prompted, select Install.

4. When the app finishes installing, it will appear on your Power BI Apps page. Select
the app and open it.
5. Select Connect your data.

6. In the Connect to Emissions Impact Dashboard for Microsoft 365 dialog, enter
your Microsoft 365 tenant ID. For help finding your tenant ID, see Find your
Microsoft 365 tenant ID.
 When done, select Next.

7. Connect your account:

For Authentication method, choose OAuth2.

For Privacy level setting for this data source, choose Organizational.
When done, select Sign in and connect.
8. Select the user account. Make sure to sign in with the credentials that have
appropriate admin access to your Microsoft 365 tenant (see list of approved roles
above).
 Wait for the view to build. This can take 24-48 hours. Refresh the semantic model
after 24 hours.

Update the app

Periodically you might receive update notifications from Appsource/Power BI about a
new version of the app. When you install the new version, the following options are
available:

Choose Update the workspace and the app, then select Install. This installs the update,
overwriting the existing/installed workspace and app. After installation is complete,
repeat steps 5-8 in the Install the app section above to re-establish the connection for
the new app version.

Issues
If there are any issues with the semantic model refresh/app update during the updating
process, validate the following steps and refresh the semantic model.

Follow the steps below to make sure your semantic model configurations are set
correctly:

1. Go to the workspace panel and open the app workspace.

2. Open the Scheduled Refresh option in the semantic model settings.

3. Open the Parameters section and configure the data source once again in the
Data Source section with the credentials with which you have access to the Tenant
ID with valid permissions, mentioned in the prerequisites section above.

4. Once the above steps are validated, go back to the app workspace and select the
Refresh option.
 5. Once the semantic model has refreshed successfully, select the Update App option
at the top-right corner of the app workspace.

Opting out
To opt out of the Emissions Impact Dashboard for Microsoft 365, go to the Opt out tab
in the app and follow the prompts. This step must be performed by a user with one of
the roles listed in the prerequisites section above.

Within 48 hours of submitting your opt-out request, go back to Power BI and delete the
Emissions Impact Dashboard for Microsoft 365 app. If you don't perform this step, and
the app does a data refresh more than 48 hours after your opt-out request, your tenant
will be opted back in to emissions processing.

If you opt out of the Emissions Impact Dashboard for Microsoft 365, you can always opt
back in by reinstalling and reconnecting to the app, but historical data won't initially be
available.

Data schemas
Data schema: GHG Preparation Report

If you choose to export data from the GHG Preparation Report tab, use the following
data schema to determine the definition of each output column.

ﾉ Expand table

Column Type Description Example

Year Integer The year in which the emissions are being reported. 2023
 Column Type Description Example

Quarter String The quarter in which the emissions are being Qtr1
reported.

Month String The month in which the emissions are being reported. January

Region String The Microsoft 365 region with which the emissions are Asia
associated. Pacific

Scope String Indicates whether the emissions value is associated Scope1

with Microsoft's Scope 1, Scope 2, or Scope 3.

Carbon Emissions Double Volume of carbon dioxide equivalent, measured in 1.019

(mtCO2e) metric tons.

Data schema: Usage Report

If you choose to export data from the GHG Preparation Report tab, use the following
data schema to determine the definition of each output column.

ﾉ Expand table

Column Type Description Example

Year Integer The year in which the active usage is being reported. 2023

Quarter String The quarter in which the active usage is being reported. Qtr1

Month String The month in which the active usage is being reported. January

Microsoft 365 Integer Count of unique active users across the Microsoft 365 apps 5000
active users currently included in the report. If a given user has usage of
multiple applications, they're only counted once.

Accessing the data via the Cloud for

Sustainability API (preview)
The same underlying data that powers the Emissions Impact Dashboard for Microsoft
365 can be accessed programmatically through the Cloud for Sustainability API
(preview). Learn more here.

Additional resources
 The role of embodied carbon in cloud emissions: Assessing the scale and sources
of Microsoft 365 emissions, and what organizations can do to help reduce them
(PDF Download) .
The carbon benefits of cloud computing: A study on the Microsoft Cloud in
partnership with WSP
Microsoft Sustainability webpage

Finding your company’s Microsoft 365 Tenant ID

Follow the steps on this page to find your tenant ID.

FAQs

App setup
I'm receiving an error at the time of connecting my data with the dashboard. What
can I do?

First, check in the Microsoft 365 Admin Center that you have one of the roles listed in
the prerequisites section above. If you don't, request this access from your
administrator. Next, ensure you're using the correct Microsoft 365 tenant ID.

I entered my tenant ID, but my company data isn't loading. What's the issue?

The Emissions Impact Dashboard for Microsoft 365 may take 24-48 hours to load your
data after completing the connection process. Return after 24 hours and select the
Refresh button in Power BI in the app workspace, as shown in the Issues section above.

How do I know that I have the latest version of the template app installed?

Microsoft might periodically release a new version of the Emissions Impact Dashboard
for Microsoft 365 in order to deliver new features or update text and visuals in the
report. Microsoft notifies your organization of new version releases via the Microsoft
365 Message center. In addition, the user in your organization who installed the
application will receive a notification in Power BI requesting that they update to the
latest version of the report. That user should follow the steps above to update to the
latest version of the app.

Emissions data
I successfully connected to my tenant's data. Why do I still see demo data in the
report, even after waiting 48 hours?

This could indicate that you don't have one of the Microsoft 365 admin roles listed in
the prerequisites section above. Make sure that one of these roles is assigned to your
profile, then open the app workspace and refresh the semantic model (see further
instructions in the Issues section above).

I successfully connected to my tenant's data. Why do I see blank data in the report?

This likely indicates that your tenant's emissions volumes are very small. The charts in
this report don't display emissions values that are lower than 0.001 MTCO2e.

Why can't I see emissions data for the previous month?

Emissions data for a given month will be available by the 14th day after the end of that
month (including nonbusiness days). Ensure that your semantic model is scheduled to
refresh automatically on a daily or weekly basis so that you always have access to the
latest information.

Why can't I see usage information prior to June 2022 in the Carbon Intensity tab?

Currently, information on Microsoft 365 usage is only available in the report from June
2022 onwards. For more usage history, see the Usage report in the Microsoft 365 Admin
Center.

Why do I see emissions from regions outside of my Microsoft 365 data location?

Microsoft 365 emissions are allocated based on data storage as well as active usage and
compute associated with Microsoft 365 applications. Emissions allocated based on
compute may occur in regions outside your Microsoft 365 data location. Learn more
about Microsoft 365 data residency here.

Where is the data used to produce this report stored?

It's stored in the United States.

How can I influence my tenant’s Microsoft 365 emissions numbers?

Refer to this white paper (PDF Download) for guidance on how to interpret and act on
the information reported in the Emissions Impact Dashboard for Microsoft 365.

Why can I only see emissions data for the past 12 months?

Currently, the Emissions Impact Dashboard for Microsoft 365 only reports on emissions
from the prior 12 months, and it isn't possible for app users to access emissions data
prior to this window. If in the future this timeframe will be extended, it will be made
available via a new app version release and announced via the Microsoft 365 Message
center.

In the 'carbon intensity' tab, why do I see my emissions grow even though the count
of Microsoft 365 users in our tenant stayed flat or shrunk (or vice versa) from one
month to the next?

It's possible for your Microsoft 365 emissions figure to move in the opposite direction of
your count of unique active users in Microsoft 365. This most often occurs when an
organization that already uses one or multiple Microsoft 365 applications begins the
process of onboarding to a new application. In this scenario, the organization would
start getting allocated emissions associated with the new application, but the overall
unique active user count displayed in the 'carbon intensity' tab of the dashboard might
remain constant or even decline. This is because the 'user count' figure only counts each
user once, even if they use multiple applications. So, for example, if an organization has
500 users of Exchange in June (and doesn't use any other Microsoft 365 applications
that month) and later onboards all 500 of those users to SharePoint in July, their overall
emissions figure would go up in July but their unique active user count would remain
the same. The Emissions Impact Dashboard only displays a single unique active user
count in the 'carbon intensity' tab; to parse out usage of specific Microsoft 365
applications, organizations can visit the Apps usage report in the Microsoft 365 Admin
Center.

Are emissions from Microsoft Copilot for Microsoft 365 usage included in the report?

Copilot for Microsoft 365 isn't currently included in this report.

Methodology
What is the methodology behind the tool?

The Emissions Impact Dashboard for Microsoft 365 reflects the specific cloud services
consumed and the associated energy requirements, efficiency of the datacenters
providing those services, electricity fuel mixes in the regions in which those datacenters
operate, and Microsoft's purchases of renewable energy. As part of the app's
development, the methodology and its implementation went through third-party
verification to ensure that it aligns to the World Resources Institute (WRI)/World
Business Council for Sustainable Development (WBCSD) Greenhouse Gas (GHG) Protocol
Corporate Accounting and Reporting Standard. The scope of the verification, conducted
in accordance with ISO 14064-3: Greenhouse gases--Part 3: Specification with guidance
for the validation and verification of greenhouse gas assertions, included the estimation
of emissions from Microsoft 365 services, but excluded the estimation of on-premises
emissions given the counterfactual nature of that estimate. A more detailed description
of the carbon calculation is documented in the Calculation Methodology tab in the
tool.

Characterizing on-premises emissions

Where does the Emissions Impact Dashboard for Microsoft 365 obtain data about my
on-premises emissions and operations?

The Emissions Impact Dashboard for Microsoft 365 doesn't obtain any information
specifically about your on-premises datacenters except what you provide. As described
in subsequent FAQs, the Emissions Impact Dashboard for Microsoft 365 relies on
industry research and user inputs about the efficiency and energy mix of on-premises
alternatives to develop an estimate of on-premises emissions.

What are the assumptions regarding on-premises estimations? Are efficiency savings
just from improvements in Power Usage Effectiveness (PUE)?

Efficiencies associated with Microsoft cloud services include far more than improved
PUE. While Microsoft datacenters strive to optimize PUE, the primary efficiency
improvements come from IT operational efficiency (dynamic provisioning, multitenancy,
server utilization) and IT equipment efficiency (tailoring hardware to services ensuring
more energy goes towards useful output), in addition to datacenter infrastructure
efficiency (PUE improvements). Our 2018 study quantifies these savings compared to
a range of on-premises alternatives ranging from low-efficiency to high-efficiency
datacenters. These findings are used to estimate the energy use required for a
corresponding on-premises datacenter to provide the same services that each customer
consumes on the Microsoft cloud.

What is the assumed energy mix for the on-premises infrastructure?

By default, the Emissions Impact Dashboard for Microsoft 365 estimates on-premises
emissions based on the mix of renewables and non-renewables on the grid. It's assumed
that the on-premises datacenter would be located on the same grid as Microsoft's
datacenters. However, for customers who purchase renewable electricity in addition
to what's on the grid (for example, through Power Purchase Agreements), users can
select the percentage of renewable electricity, and the Emissions Impact Dashboard for
Microsoft 365 will adjust on-premises emissions accordingly.

When should I choose Low, Medium, or High for the efficiency of the on-premises
infrastructure?
Users should select the efficiency most representative of the on-premises deployment
they would like to compare against, based on the equipment and datacenter
characteristics here:

Low: Physical servers and direct attached storage in small localized datacenter
(500-1,999 square feet).
Medium: Mix of physical/virtualized servers and attached/dedicated storage in
mid-tier internal datacenter (2,000-19,999 square feet).
High: Virtualized servers and dedicated storage in high-end internal datacenter
(>20,000 square feet).
Connect to Office365Mon with Power BI
Article • 11/10/2023

Analyzing your Office 365 outages and health performance data is easy with Power BI
and the Office365Mon template app. Power BI retrieves your data, including outages
and health probes, and then builds an out-of-the-box dashboard and reports based on
that data.

Connect to the Office365Mon template app for Power BI.

７ Note

You need an Office365Mon admin account to connect to and load the Power BI
template app.

How to connect
1. Select Connect your data at the top of the screen:

2. In the Connect to Office365 Power BI Template Pack window, select Next:

3. In the Authentication method box, select OAuth2. You can change the privacy
level if you want. For more information, select Learn more in the window. When
you're done, select Sign in and connect.
4. When prompted, enter your Office365Mon admin credentials and complete the
authentication process.

5. After Power BI imports the data, you see a new dashboard, report, and semantic
model in your workspace. Select Office365Mon.
 

What now?

Try asking a question in the Q&A box at the top of the dashboard.
Change the tiles in the dashboard.
Select a tile to open the underlying report.
Change the refresh schedule. The semantic model is scheduled to refresh daily. You
can change the schedule, or refresh it on demand by selecting Refresh now in the
workspace.

Troubleshooting
If you see a Need admin approval error when you try to sign in to Office365Mon, the
account you're using doesn't have permissions to retrieve the data. You need to use an
Office365Mon admin account when you sign in.

Next steps
What is Power BI?
Get data for Power BI
Connect to Project Web App with Power
BI
Article • 11/10/2023

Microsoft Project Web App is a flexible online solution for project portfolio management
(PPM) and everyday work. Project Web App enables organizations to get started,
prioritize project portfolio investments and deliver the intended business value. The
Project Web App Template App for Power BI allows you to unlock insight from Project
Web App to help manage projects, portfolios and resources.

Connect to the Project Web App Template App for Power BI.

How to connect
1. Select Apps in the nav pane > select Get apps in the upper right corner.

2. In the Services box, select Get.

3. In AppSource, select the Apps tab, and search/select Microsoft Project Web App.

4. You will get a message saying - Install this Power BI App? select Install.
5. In the Apps pane, select the Microsoft Project Web App tile.

6. In Get started with your new app, select Connect data.

7. In the Project Web App URL text box, enter the URL for the Project Web App
(PWA) you want to connect to. Note this may differ from the example if you have a
custom domain. In the PWA Site Language text box, type the number that
corresponds to your PWA site language. Type the single digit '1' for English, '2' for
French, '3' for German, '4' for Portuguese (Brazil), '5' for Portuguese (Portugal) and
'6' for Spanish.
 8. For Authentication Method, select oAuth2 > Sign In. When prompted, enter your
Project Web App credentials and follow the authentication process.

７ Note

You need to have Portfolio Viewer, Portfolio Manager, or Administrator

permissions for the Project Web App you are connecting to.

9. You'll see a notification indicating your data is loading. Depending on the size of
your account this may take some time. After Power BI imports the data, you will
see the contents of your new workspace. You may need to refresh the semantic
model to get the latest updates.

After Power BI imports the data you will see the report with 13 pages and semantic
model in the nav pane.

10. Once your reports are ready, go ahead and start exploring your Project Web App
data! The Template App comes with 13 rich and detailed reports for the Portfolio
Overview (6 report pages), Resource Overview (5 report pages) and Project Status
(2 report pages).
What now?

While your semantic model will be scheduled to refresh daily, you can change the
refresh schedule or try refreshing it on demand using Refresh Now.

Expand the Template App

Download the GitHub PBIT file to further customize and update the template app.

Next steps
Get started in Power BI

Get data in Power BI

Connect to the Regional Emergency
Response Dashboard
Article • 01/23/2024

The Regional Emergency Response Dashboard is the reporting component of the

Microsoft Power Platform Regional Emergency Response solution. Regional organization
admins can view the dashboard in their Power BI tenant, enabling them to quickly view
important data and metrics that will help them make efficient decisions.

This article tells you how to install the Regional Emergency Response app using the
Regional Emergency Response Dashboard template app, and how to connect to the
data sources.

For detailed information about what is presented in the dashboard, see Get insights.

After you've installed the template app and connected to the data sources, you can
customize the report as per your needs. You can then distribute it as an app to
colleagues in your organization.

Prerequisites
Before installing this template app, you must first install and set up the Regional
Emergency Response solution. Installing this solution creates the datasource references
necessary to populate the app with data.

When installing Regional Emergency Response solution, take note of the URL of your
Common Data Service environment instance. You will need it to connect the template
app to the data.

Install the app

1. Click the following link to get to the app: Regional Emergency Response
Dashboard template app

2. On the AppSource page for the app, select GET IT NOW .

3. Select Install.

Once the app has installed, you see it on your Apps page.
Connect to data sources
1. Select the icon on your Apps page to open the app.

2. On the splash screen, select Explore.

The app opens, showing sample data.

3. Select the Connect your data link on the banner at the top of the page.
4. In the dialog box that appears, type the URL of your Common Data Service
environment instance. For example: https://[myenv].crm.dynamics.com. When
done, click Next.

5. In the next dialog that appears, set the authentication method to OAuth2. You
don't have to do anything to the privacy level setting.

Select Sign in.

6. At the Microsoft sign-in screen, sign in to Power BI.

After you've signed in, the report connects to the data sources and is populated
with up-to-date data. During this time, the activity monitor turns.
Schedule report refresh
When the data refresh has completed, set up a refresh schedule to keep the report data
up to date.

1. In the top header bar, select Power BI.

2. In the left navigation pane, look for the Regional Emergency Response Dashboard
workspace under Workspaces, and follow the instructions described in the
Configure scheduled refresh article.

Customize and share

See Customize and share the app for details. Be sure to review the report disclaimers
before publishing or distributing the app.

Related content
Understanding the Regional Emergency Response dashboard
Set up and learn about the Crisis Communication sample template in Power Apps
Questions? Try asking the Power BI Community
What are Power BI template apps?
Install and distribute template apps in your organization
Salesforce Analytics for Sales Managers
Article • 03/07/2022

The Salesforce Analytics for Sales Managers template app includes visuals and insights
for analyzing your marketing effort.

The app's out-of-the-box dashboard provides key metrics, such as your sales pipeline,
best accounts, and KPI's. You can drill down into the report for more details on each
aspect. Fully interactive visuals help you explore your data further.

In this article we will walk through the app using sample data to give you an idea of how
you can use the app to gain key insights into your sales data.

Prerequisites
Power BI Pro
A marketing, developer, or admin Salesforce subscription

Install the app

1. Click the following link to get to the app: Salesforce Analytics for Sales Managers
template app.

2. Once you're on the App's AppSource page, click GET IT NOW .

3. When prompted , click Install. Once the app has installed, you will see it on your
Apps page.
Connect to data sources
1. Click the icon on your Apps page to open the app. The app opens, showing sample
data.

2. Select the Connect your data link on the banner at the top of the page.

3. The parameters dialog will appear. There are no required parameters. Click Next.
4. The authentication method dialog will appear. Recommended values are
prepopulated. Don't change these unless you have specific knowledge of different
values. Click Sign in and connect.
 5. When prompted, sign into Salesforce.

The report will connect to the data sources and be populated with up-to-date
data. During this time you will see sample data and that refresh is in progress.

What does the KPI dashboard tell us?

When you open the app, you'll see the KPI dashboard. The KPI dashboard shows us an
overall view of the key metrics from all the dashboards. Click the arrows to get to the
individual dashboards.
What does the Sales Manager dashboard tell
us?
The Sales Manager dashboard and underlying report focus on a typical sales challenge:
providing a total sales analysis over a certain period.

Using the dashboard to look at the sample data, we're going see how to

Understand how much we can generate from the total open opportunities that a
Sales company has.
Identify which areas we should focus on in the total sales life cycle.
Spot loopholes we should look into.

Now let's look at the various components of the dashboard.

The top two visuals show us the total number of open opportunities we have and the
revenue we can expect to generate from them.

The Opportunity Sales Stage visual shows the position of the opportunities in the sales
pipeline. Select any stage on the sales pipeline stage to see its impact on the whole
sales process. Now we can analyze the revenue that corresponds with that stage.
If you click on any one of the industries in the Share of Industry visual, it shows the share
of the industry your sales team is working on. Let's click on "Apparel" for example.

When we click on "Apparel", we see that the account we have in this sector is one of the
top ten critical opportunities with respect to revenue. We can conclude that this is an
opportunity that deserves our focus on a priority basis, since its expected revenue is
among the highest as compared to other accounts.
What does the Account dashboard tell us?
The Account dashboard lets you oversee how you are performing in all your accounts. It
tells you which are your most profitable accounts. Let's analyze the accounts for apparel.

On the dashboard, select “Apparel” on the Account dashboard's Account Share Industry
Wise visual.

You'll see the tiles get updated. Notice that there is one industry in apparel, for which we
have 1 account.

If we look at the Account Area Wise map visual, you'll see the area where we have the
apparel account.
Let’s look at the revenue for the account under apparel. In the Revenue by Account
visual you can see that the account for apparel is highlighted, and shows the revenue
being generated.

We can also see a comparison of revenue won vs lost. If we hover over a bar, we see the
exact revenue won out of the total revenue.
What does the Lead dashboard tell us?
The Lead dashboard lets you see what the sources of your leads are. It tells you which
are your most profitable sources of lead.

Look at the Probability of Conversion visual to examine what the probability is that the
lead from a source is going to be converted. For example, selecting "External Referral" in
the Leads by Source visual shows that the probability of conversion is 90.00%, while its
total forecasted amount is 1.65 million dollars.

Similarly, you can also see the distribution of customers in the sales lead by looking at
the Sales Lead by Customer Type visual. You can see it for a single lead by hovering over
it. So, for "External Referral", you can see that there is one new customer belonging to
the “New Business” category.

We can also see the overall number of lead statuses that we have, and in what stage
they are.
What does Representative dashboard tell us?
The Representative dashboard lets you measure the performance of the sales
representatives by a number of matrices.

We can see all the industries a sales representative works in, and all the representatives
of an industry.

To see the performance of the sales representatives, we selected "Energy" as the

industry from the Share of Industry visual. Then, in the Total Opportunities Gained visual,
we can see the name of the sales representative and the number of opportunities which
belong to the selected industry.

The top three visuals show us the number of leads generated by the sales
representatives and the revenue they have generated and lost.

At same time, we can use the Month Wise Performance visual to measure the month-
wise performance of the sales representatives.
System requirements and considerations
Connected with a production Salesforce account that has API access enabled.
Permission granted to the Power BI app during sign in.
The account has sufficient API calls available to pull and refresh the data.
A valid authentication token is required for refresh. Salesforce has a limit of five
authentication tokens per application so make sure you've five or less Salesforce
data sets imported.
The Salesforce Reports API has a restriction that supports up to 2,000 rows of data.

Troubleshooting
If you encounter any errors, review the requirements above.
Signing in to a custom or sandbox domain isn't currently supported.
Salesforce connector reference

"Unable to connect to the remote server" message

If you get an "Unable to connect to the remote server" message when trying to connect
to your Salesforce account, see this solution on the following forum: Salesforce
Connector sign in Error Message: Unable to connect to the remote server

Next steps
What are Power BI template apps
Create a template app in Power BI
Install and distribute template apps in your organization
Questions? Try asking the Power BI Community
Connect to Smartsheet with Power BI
Article • 11/10/2023

This article walks you through pulling your data from your Smartsheet account with a
Power BI template app. Smartsheet offers an easy platform for collaboration and file
sharing. The Smartsheet template app for Power BI provides a dashboard, reports, and
semantic model that show an overview of your Smartsheet account. You can also use
Power BI Desktop to connect directly to individual sheets in your account.

After you've installed the template app, you can change the dashboard and report. Then
you can distribute it as an app to colleagues in your organization.

Connect to the Smartsheet template app for Power BI.

７ Note

A Smartsheet admin account is preferred for connecting and loading the Power BI
template app as it has additional access.

Install the app

1. Select Apps in the navigation pane, then choose Get apps in the upper-right
corner.

2. In Power BI apps, select the Apps tab, and search for the service you want.
 1. Select Smartsheet > Get it now.

2. In Install this Power BI App? select Install.

3. In the Apps pane, select the Smartsheet tile.

Connect to your Smartsheet data source

1. Select the Smartsheet tile on your Apps page to open the app. The app opens,
showing sample data.

2. Select the Connect your data link on the banner at the top of the page.

3. For Authentication Method, select oAuth2 > Sign In.

 When prompted, enter your Smartsheet credentials and follow the authentication
process.

4. After Power BI imports the data, the Smartsheet dashboard opens.

Modify and distribute your app
You've installed the Smartsheet template app. That means you've also created the
Smartsheet workspace. In the workspace, you can change the report and dashboard,
and then distribute it as an app to colleagues in your organization.

1. To view all the contents of your new Smartsheet workspace, in the nav pane, select
Workspaces > Smartsheet.

This view is the content list for the workspace. In the upper-right corner, you see
Update app. When you're ready to distribute your app to your colleagues, that's
where you'll start.
 2. Select Reports and Semantic models to see the other elements in the workspace.

Read about distributing apps to your colleagues.

What's included
The Smartsheet template app for Power BI includes an overview of your Smartsheet
account, such as the number of workspaces, reports, and sheets you have, when they're
modified etc. Admin users also see some information around the users in their system,
such as top sheet creators.

To connect directly to individual sheets in your account, you can use the Smartsheet
connector in the Power BI Desktop.

Next steps
Create workspaces in Power BI
Install and use apps in Power BI
Connect to Power BI apps for external services
Questions? Try asking the Power BI Community
Connect to Zendesk with Power BI
Article • 07/25/2023

This article walks you through pulling your data from your Zendesk account with a
Power BI template app. The Zendesk app offers a Power BI dashboard and a set of
Power BI reports that provide insights about your ticket volumes and agent
performance. The data is refreshed automatically once a day.

After you've installed the template app, you can customize the dashboard and report to
highlight the information you care about most. Then you can distribute it as an app to
colleagues in your organization.

Connect to the Zendesk template app or read more about the Zendesk integration
with Power BI.

After you've installed the template app, you can change the dashboard and report. Then
you can distribute it as an app to colleagues in your organization.

７ Note

You need a Zendesk Admin account to connect. More details on requirements

below.

２ Warning

Before Oct 15, 2019, the Zendesk Support Search API allowed for a total of 200,000
results to be received through pagination of large queries. To align search usage
with its intended scope, Zendesk now limits the maximum number of results
returned to 1,000 total results, with a maximum of 100 results per page. However,
the current Power BI Zendesk connector can still create API calls that exceed these
new limits, resulting in possibly misleading results.

Install the app

1. Select Apps in the navigation pane, then choose Get apps in the upper-right
corner.
 2. In Power BI apps, select the Apps tab, and search for the service you want.

1. Select Zendesk > Get it now.

2. When prompted, select Install. Once the app has installed, you'll see it listed on
your Apps page.

Connect to your Zendesk data source

1. Select the Zendesk tile on your Apps page to open the app. The app opens,
showing sample data.

2. Select the Connect your data link on the banner at the top of the page.

3. Provide the URL associated with your account. The URL has the form
https://company.zendesk.com . See details on finding these parameters below.
4. When prompted, enter your Zendesk credentials. Select oAuth 2 as the
Authentication Mechanism and select Sign In. Follow the Zendesk authentication
flow. (If you're already signed in to Zendesk in your browser, you may not be
prompted for credentials.)

７ Note

This template app requires that you connect with a Zendesk Admin account.
5. Select Allow to allow Power BI to access your Zendesk data.

6. Select Connect to begin the import process.

 7. After Power BI imports the data, you see the content list for your Zendesk app: a
new dashboard, report, and dataset.

8. Select the dashboard to start the exploration process.

Modify and distribute your app

You've installed the Zendesk template app. That means you've also created the Zendesk
workspace. In the workspace, you can change the report and dashboard, and then
distribute it as an app to colleagues in your organization.

1. To view all the contents of your new Zendesk workspace, in the nav pane, select
Workspaces > Zendesk.
 This view is the content list for the workspace. In the upper-right corner, you see
Update app. When you're ready to distribute your app to your colleagues, that's
where you'll start.

2. Select Reports and Datasets to see the other elements in the workspace.

Read about distributing apps to your colleagues.

System requirements
A Zendesk Administrator account is required to access the Zendesk template app. If
you're an agent or an end user and are interested in viewing your Zendesk data, add a
suggestion and review the Zendesk connector in the Power BI Desktop.

Finding parameters
Your Zendesk URL will be the same as the URL you use to sign into your Zendesk
account. If you're not sure of your Zendesk URL, you can use the Zendesk login help .
Troubleshooting
If you're having issues connecting, check your Zendesk URL and confirm you're using a
Zendesk administrator account.

Next steps
Create workspaces in Power BI
Install and use apps in Power BI
Connect to Power BI apps for external services
Questions? Try asking the Power BI Community
Guidance for deploying a data gateway
for the Power BI service
Article • 05/23/2024

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

This article provides guidance and considerations for deploying a data gateway for the
Power BI service in your network environment.

For information about how to download, install, configure, and manage the on-premises
data gateway, see What is an on-premises data gateway?. You can also find out more
about the on-premises data gateway and Power BI by visiting the Microsoft Power BI
blog and the Microsoft Power BI Community site.

Installation considerations for the on-premises

data gateway
Before you install the on-premises data gateway for your Power BI cloud service, there
are some considerations to keep in mind. The following sections describe these
considerations.

Number of users
The number of users who consume a report that uses the gateway is an important
metric in your decision about where to install the gateway. Here are some questions to
consider:

Do users use these reports at different times of the day?

What types of connections do they use: DirectQuery or Import?
Do all users use the same report?

If all the users access a given report at the same time each day, make sure that you
install the gateway on a machine that's capable of handling all those requests. See the
following sections for performance counters and minimum requirements that can help
you determine whether a machine is adequate.

A constraint in the Power BI service allows only one gateway per report. Even if a report
is based on multiple data sources, all such data sources must go through a single
gateway. If a dashboard is based on multiple reports, you can use a dedicated gateway
for each contributing report. In this way, you distribute the gateway load among the
multiple reports that contribute to the single dashboard.

Connection type
The Power BI service offers two types of connections: DirectQuery and Import. Not all
data sources support both connection types. Many factors might contribute to your
choice of one over the other, such as security requirements, performance, data limits,
and data model sizes. To learn more about connection types and supported data
sources, see the list of available data source types.

Depending on which type of connection is used, gateway usage can be different. For
example, try to separate DirectQuery data sources from scheduled refresh data sources
whenever possible. The assumption is that they're in different reports and can be
separated. Separating sources prevents the gateway from having thousands of
DirectQuery requests queued up at the same time as the morning's scheduled refresh of
a large-size data model that's used for the company's main dashboard.

Here's what to consider for each option:

Scheduled refresh: Depending on your query size and the number of refreshes
that occur per day, you can choose to stay with the recommended minimum
hardware requirements or upgrade to a higher performance machine. If a given
query isn't folded, transformations occur on the gateway machine. As a result, the
gateway machine benefits from having more available RAM.

DirectQuery: A query is sent each time any user opens the report or looks at data.
If you expect more than 1,000 users to access the data concurrently, make sure
your computer has robust and capable hardware components. More CPU cores
result in better throughput for a DirectQuery connection.

For the machine installation requirements, see the on-premises data gateway
installation requirements.

Location
The location of the gateway installation can have significant effect on your query
performance. Try to make sure that your gateway, data source locations, and the Power
BI tenant are as close as possible to each other to minimize network latency. To
determine your Power BI tenant location, in the Power BI service select the question
mark (?) icon in the upper-right corner. Then select About Power BI.

If you intend to use the Power BI service gateway with Azure Analysis Services, be sure
that the data regions in both match. For more information about how to set data
regions for multiple services, watch this video .

Optimizing performance
By default, the gateway spools data before returning it to the dataset, potentially
causing slower performance during data load and refresh operations. The default
behavior can be overridden.

1. In the C:\Program Files\On-Premises data

gateway\Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll.config file,
set the StreamBeforeRequestCompletes property to True , and then save.

JSON

<setting name="StreamBeforeRequestCompletes" serializeAs="String">

<value>True</value>
</setting>
 2. In On-premises data gateway > Service Settings, restart the gateway.

If installing the gateway on an Azure Virtual Machine, ensure optimal networking

performance by configuring accelerated networking. To learn more, see Create a
Windows VM with accelerated networking.

Related content
Configure proxy settings
Troubleshoot gateways - Power BI
On-premises data gateway FAQ - Power BI

More questions? Try the Power BI Community .

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

On-premises data gateway in-depth
Article • 01/23/2024

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

We moved the information from this article to several articles across the Power BI and
general docs. Follow the links under each heading to find the relevant content.

How the gateway works

See On-premises data gateway architecture.

List of available data source types

See Add or remove a gateway data source.

Authentication to on-premises data sources

See Authentication to on-premises data sources.

Authentication to a live Analysis Services data

source
See Authentication to a live Analysis Services data source.

Role-based security
See Role-based security.

Row-level security
See Row-level security.

What about Microsoft Entra ID?

See Microsoft Entra ID.

How do I tell what my UPN is?

See How do I tell what my UPN is?.

Map user names for Analysis Services data

sources
See Map user names for Analysis Services data sources.

Synchronize an on-premises Active Directory

with Microsoft Entra ID
See Synchronize an on-premises Active Directory with Microsoft Entra ID.

What to do next?
See the articles on data sources:

Add or remove a gateway data source

Manage your data source - Analysis Services
Manage your data source - SAP HANA
Manage your data source - SQL Server
Manage your data source - Oracle
Manage your data source - Import/Scheduled refresh

Where things can go wrong

See Troubleshoot the on-premises data gateway and Troubleshoot gateways - Power BI.

Sign in account
See Sign in account.
Windows Service account
See Change the on-premises data gateway service account.

Ports
See Ports.

Forcing HTTPS communication with Azure

Service Bus
See Force HTTPS communication with Azure Service Bus.

Support for TLS 1.2

See TLS 1.2 for gateway traffic.

How to restart the gateway

See Restart a gateway.

Related content
What is the on-premises data gateway?
More questions? Try the Power BI Community
Use a personal gateway in Power BI
Article • 05/28/2024

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

The on-premises data gateway (personal mode) is a version of the on-premises data
gateway that works only with Power BI. You can use a personal gateway to install a
gateway on your own computer and get access to on-premises data.

７ Note

Each Power BI user can have only one personal mode gateway running. If the same
user installs another personal mode gateway, even on a different computer, the
most recent installation replaces the existing previous installation.

On-premises data gateway vs. on-premises

data gateway (personal mode)
The following table describes differences between an on-premises data gateway and an
on-premises data gateway (personal mode).

ﾉ Expand table

On-premises data gateway On-premises data gateway

(personal mode)

Supports cloud Power BI, PowerApps, Azure None

services: Logic Apps, Power Automate,
Azure Analysis Services,
dataflows

Runs under As configured by users who Your credentials for Windows

credentials: have access to the gateway authentication, or credentials you
configure for other authentication
types
 On-premises data gateway On-premises data gateway
(personal mode)

Can install only as Yes No

computer admin

Centralized gateway Yes No

and data source
management

Can import data and Yes Yes

schedule refresh

DirectQuery support Yes No

LiveConnect support Yes No

for Analysis Services

Install the on-premises data gateway (personal

mode)
To install the on-premises data gateway (personal mode):

1. Download the on-premises data gateway .

2. Open the installer, and select Next.

3. Select On-premises data gateway (personal mode), and then select Next.
 4. On the next screen, review the minimum requirements, verify or edit the
installation path, and select the checkbox to accept the terms of use and privacy
statement. Then select Install.

5. After the installation completes successfully, enter your email address under Email
address to use with this gateway, and select Sign in.

6. After you sign in, a confirmation screen displays.

7. Select Close to close the installer.

Use Fast Combine with the personal gateway

Fast Combine on a personal gateway helps you ignore specified privacy levels when you
run queries. To enable Fast Combine for the on-premises data gateway (personal mode):

1. Use Windows File Explorer to open the file <localappdata>\Microsoft\On-premises

data gateway (personal
mode)\Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll.config.

2. At the end of the file, before

</Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.GatewayCoreSettings> ,

add the following code, and then save the file.

 XML

<setting name="EnableFastCombine" serializeAs="String">

<value>true</value>
</setting>

3. The setting takes effect in approximately one minute. To confirm that Fast Combine
is working properly, try an on-demand refresh in the Power BI service.

Frequently asked questions (FAQ)

Question: Can you run the on-premises data gateway (personal mode) side-by-
side with the on-premises data gateway that used to be called the Enterprise
gateway?

Answer: Yes, both gateways can run simultaneously.

Question: Can you run the on-premises data gateway (personal mode) as a
service?

Answer: No. The on-premises data gateway (personal mode) can run only as an
application. To run a gateway as a service or in admin mode, use the on-premises
data gateway, which used to be called the Enterprise gateway.

Question: How often does the on-premises data gateway (personal mode) update?

Answer: The personal gateway updates monthly.

Question: Why does the personal gateway ask you to update your credentials?

Answer: Many situations can trigger a request for credentials. The most common
scenario is that you reinstalled the on-premises data gateway (personal mode) on
a different machine than your original Power BI personal gateway. There could also
be an issue in the data source, or Power BI failed to make a test connection, or a
timeout or system error occurred.

To update your credentials in the Power BI service, select the gear icon in the
header and then choose Settings. On the Semantic models tab, select the dataset,
and then choose Data source credentials.

Question: How long is a personal gateway offline during an upgrade?

Answer: Upgrading the personal gateway to a new version takes only few minutes.

Question: Does the personal gateway support R and Python scripts?

 Answer: Yes, personal mode supports R and Python scripts.​

Related content
Add or remove a gateway data source
Configure proxy settings for an on-premises data gateway
Power BI implementation planning: Data gateways

More questions? Try the Power BI Community .

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

On-premises data gateway FAQ -
Power BI
FAQ

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

Do I need to upgrade the on-premises

data gateway (personal mode)?
No, you can keep using the on-premises data gateway (personal mode) for Power BI.

Are any special permissions required to

install the gateway and manage it in the
Power BI service?
No special permissions are required. You need to sign in with either a work or school
email account.

Can I upload Excel workbooks with

Power Pivot data models that connect
to on-premises data sources, and do I
need a gateway for this scenario?
Yes, you can upload the workbook. No, you don’t need a gateway. But, because the data
will reside in the Excel data model, reports in Power BI based on the Excel workbook
won't be live. To refresh reports in Power BI, you have to reupload an updated workbook
each time. Or, use the gateway with scheduled refresh.
If users share dashboards with a
DirectQuery connection, will other users
see the data even though they might
not have the same permissions?
For a dashboard connected to Analysis Services, users will see only the data they have
access to. If the users don't have the same permissions, they won't be able to see any
data. For other data sources, all users will share the credentials entered by the admin for
that data source.

Why can't I connect to my Oracle

server?
You might need to install the Oracle client and configure the tnsnames.ora file with the
proper server information to connect to your Oracle server. The oracle client is a
separate installation outside of the gateway. For more information, see Install the Oracle
client.

Are R scripts supported?

R scripts are supported only for personal mode.​

Can I use msmdpump.dll to create

custom effective username mappings
for Analysis Services?
No. This use isn't supported.

Can I use the gateway to connect to a

multidimensional (OLAP) instance?
Yes. The on-premises data gateway supports live connections to both Analysis Services
Tabular and Multidimensional models.
What if I install the gateway on a
computer in a different domain from
my on-premises server that uses
Windows authentication?
No guarantees. It depends on the trust relationship between the two domains. If two
different domains are in a trusted domain model, the gateway might be able to connect
to the Analysis Services server, and the effective username can be resolved. If not, you
might encounter a sign-in failure.

How can I find out what effective

username is being passed to my on-
premises Analysis Services server?
See Troubleshoot gateways - Power BI.

Next steps
Troubleshoot the on-premises data gateway
Power BI implementation planning: Data gateways

More questions? Ask the Power BI Community .

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Add or remove a gateway data source
Article • 12/21/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

Power BI supports many on-premises data sources, and each source has its own
requirements. You can use a gateway for a single data source or multiple data sources.
For this example, you learn how to add SQL Server as a data source. The steps are
similar for other data sources.

You can also do most data sources management operations by using APIs. For more
information, see REST APIs (Gateways).

If you don't have a gateway installed, install an on-premises data gateway to get started.

Add a data source

1. From the page header in the Power BI service, select the Settings icon, and then
select Manage connections and gateways.

2. Select New at the top of the screen to add a new data source.
3. On the New connection screen, select On-premises, provide the Gateway cluster
name you want to create the connection on, provide a Connection name, and
select the Data Source Type. For this example, choose SQL Server.

4. Enter information about the data source. For SQL Server, provide the Server and
Database.

７ Note

To use the data source for Power BI reports and dashboards, the server and
database names must match between Power BI Desktop and the data source
you add to the gateway.

5. Select an Authentication Method to use when connecting to the data source,

Basic, Windows, or OAuth2. For SQL Server, choose Windows or Basic (SQL
Authentication). Enter the credentials for your data source.
 If you selected OAuth2 authentication method:

Any query that runs longer than the OAuth token expiration policy may fail.
Cross-tenant Microsoft Entra accounts aren't supported.

If you selected Windows authentication method, make sure that account has
access on the machine. If you're not sure, make sure to add NT-
AUTHORITY\Authenticated Users (S-1-5-11) to the local machine Users group.

6. Optionally, under Single sign-on, you can configure single sign-on (SSO) for your
data source.

Depending on your organization settings, for DirectQuery-based reports, you can

configure Use SSO via Kerberos for DirectQuery queries, Use SSO via Kerberos
for DirectQuery And Import queries or Use SSO via Microsoft Entra ID for
DirectQuery queries. You can configure Use SSO via Kerberos for DirectQuery
And Import queries for refresh-based reports.

If you use Use SSO via Kerberos for DirectQuery queries and use this data source
for a DirectQuery-based report, the report uses the credentials of the user that
signs in to the Power BI service. A refresh-based report uses the credentials that
you enter in the Username and Password fields and the Authentication method
you choose.
 When you use Use SSO via Kerberos for DirectQuery And Import queries, you
don't need to provide any credentials. If this data source is used for DirectQuery-
based reports, the report uses the user that's mapped to the Microsoft Entra user
that signs in to the Power BI service. A refresh-based report uses the dataset
owner's security context.

For more information about Use SSO via Kerberos for DirectQuery queries and
Use SSO via Kerberos for DirectQuery And Import queries, see Overview of single
sign-on (SSO) for gateways in Power BI.

If you use Use SSO via Microsoft Entra ID for DirectQuery queries and use this
data source for a DirectQuery-based report, the report uses the Microsoft Entra
token of the user who signs into the Power BI service. A refresh-based report uses
the credentials that you enter in the Username and Password fields and the
Authentication method you choose. The Use SSO via Microsoft Entra ID for
DirectQuery queries option is available only if the tenant admin allows Microsoft
Entra SSO via the on-premises data gateway, and for the following data sources:

SQL Server
Azure Data Explorer
Snowflake

For more information about Use SSO via Microsoft Entra ID for DirectQuery
queries, see Microsoft Entra single sign-on (SSO) for Gateway.

７ Note

SSO for Import queries is available only for the SSO data sources that use
Kerberos constrained delegation.

7. Under General > Privacy level, optionally configure a privacy level for your data
source. This setting doesn't apply to DirectQuery.
 8. Select Create. Under Settings, you see Created new connection if the process
succeeds.

You can now use this data source to include data from SQL Server in your Power BI
dashboards and reports.

Remove a data source

You can remove a data source if you no longer use it. If you remove a data source, any
dashboards and reports that rely on that data source no longer work.

To remove a data source, select the data source from the Data (preview) screen in
Manage connections and gateways, and then select Remove from the top ribbon.

Use the data source for scheduled refresh or

DirectQuery
After you create the data source, it's available to use with DirectQuery connections or
through scheduled refresh. You can learn more about setting up scheduled refresh in
Configure scheduled refresh.

The link between your dataset and the data source in the gateway is based on your
server name and database name. These names must match. For example, if you supply
an IP address for the server name in Power BI Desktop, you must use the IP address for
the data source in the gateway configuration. If you use SERVER\INSTANCE in Power BI
Desktop, you must use the same format in the data source you configure for the
gateway.

If you're listed in the Users tab of the data source configured in the gateway, and the
server and database name match, you see the gateway listed as Running under
Gateway connections in the Settings for your data source. You can select Scheduled
refresh to set up scheduled refresh for the data source.

） Important

If your dataset contains multiple data sources, each data source must be added in
the gateway. If one or more data sources aren't added to the gateway, you won't
see the gateway as available for scheduled refresh.

Manage users
After you add a data source to a gateway, you give users and security groups access to
the specific data source, not the entire gateway. The access list for the data source
controls only who is allowed to publish reports that include data from the data source.
Report owners can create dashboards and apps, and then share those items with other
users.
You can also give users and security groups administrative access to the gateway.

７ Note

Users with access to the data source can associate datasets to the data source, and
connect, based on either the stored credentials or SSO you selected while creating
a data source.

Add users to a data source

1. From the page header in the Power BI service, select the Settings icon, and then
select Manage connections and gateways.

2. Select the data source where you want to add users.

3. Select Manage users from the top ribbon

4. On the Manage users screen, enter the users and/or security groups from your
organization who can access the selected data source.

5. Select the new user name, and select the role to assign: User, User with resharing,
or Owner.

6. Select Share, and the added member's name is added to the list of people who can
publish reports that use this data source.

Remember that you need to add users to each data source that you want to grant
access to. Each data source has a separate list of users. Add users to each data source
separately.
Remove users from a data source
On the Manage Users tab for the data source, you can remove users and security
groups that use this data source.

Store encrypted credentials in the cloud

When you add a data source to the gateway, you must provide credentials for that data
source. All queries to the data source run by using these credentials. The credentials are
encrypted securely with symmetric encryption, so that they can't be decrypted in the
cloud. The credentials are sent to the machine that runs the on-premises gateway,
where they're decrypted when the data sources are accessed.

List of available data source types

For information about which data sources the on-premises data gateway supports, see
Power BI data sources.

Next steps
Manage your data source - Analysis Services
Manage your data source - SAP HANA
Manage your data source - SQL Server
Manage your data source - Oracle
Manage your data source - Import/scheduled refresh
Guidance for deploying a data gateway

More questions? Try the Power BI Community .

Manage SQL Server Analysis Services
data sources
Article • 12/21/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

After you install an on-premises data gateway, you can add data sources to use with the
gateway. This article describes how to add a SQL Server Analysis Services (SSAS) data
source to your on-premises gateway to use for scheduled refresh or for live connections.

To learn more about how to set up a live connection to SSAS, watch this Power BI
Walkthrough: Analysis Services Live Connect video.

７ Note

If you have an Analysis Services data source, you need to install the gateway on a
computer joined to the same forest or domain as your Analysis Services server.

７ Note

The gateway supports only Windows authentication for Analysis Services.

Add a data source

To connect to either a multidimensional or tabular Analysis Services data source:

1. On the New connection screen for your on-premises data gateway, select Analysis
Services for Connection type. For more information about how to add a data
source, see Add a data source.
2. Fill in the information for the data source, which includes Server and Database.
The gateway uses the information you enter for Username and Password to
connect to the Analysis Services instance.

７ Note

The Windows account you enter must be a member of the Server

Administrator role on the Analysis Services instance you're connecting to. If
 this account's password is set to expire, users get a connection error unless
you update the data source password. For more information about how
credentials are stored, see Store encrypted credentials in the cloud.

3. Configure the Privacy level for your data source. This setting controls how data can
be combined for scheduled refresh. The privacy-level setting doesn't apply to live
connections. To learn more about privacy levels for your data source, see Set
privacy levels (Power Query) .

4. Optionally, you can configure user name mapping now. For instructions, see
Manual user name remapping.

5. After you complete all the fields, select Create.

You can now use this data source for scheduled refresh or live connections against an
on-premises Analysis Services instance.

User names for Analysis Services

To learn about authentication with Analysis Services live connections in Power BI, watch
this video:

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI service.

https://www.youtube-nocookie.com/embed/Qb5EEjkHoLg

Each time a user interacts with a report connected to Analysis Services, the effective user
name passes to the gateway and then passes on to your on-premises Analysis Services
server. The email address that you use to sign in to Power BI passes to Analysis Services
as the effective user in the EffectiveUserName connection property.

The email address must match a defined user principal name (UPN) within the local
Active Directory (AD) domain. The UPN is a property of an AD account. The Windows
account must be present in an Analysis Services role. If a match can't be found in AD,
the sign-in isn't successful. To learn more about AD and user naming, see User naming
attributes.

Map user names for Analysis Services data

sources
You can also map your Power BI sign-in name to a local directory UPN. To learn about
UPN mapping in Power BI, watch this video:

７ Note

This video might use earlier versions of Power BI Desktop or the Power BI service.

https://www.youtube-nocookie.com/embed/eATPS-c7YRU

Power BI allows mapping user names for Analysis Services data sources. You can
configure rules to map a Power BI sign-in user name to an EffectiveUserName that
passes to the Analysis Services connection. This feature is a great workaround when
your Microsoft Entra user name doesn't match a UPN in your local Active Directory
instance. For example, if your email address is [email protected] , you can
map it to [email protected] , and that value passes on to the gateway.

You can map user names for Analysis Services in two different ways:

Manual user remapping in Power BI

Active Directory lookup mapping, which uses on-premises AD property lookup to
remap Microsoft Entra UPNs to on-premises AD users.

Manual mapping by using on-premises AD property lookup is possible, but is time

consuming and difficult to maintain, especially when pattern matching isn't enough. For
example, domain names or user account names might be different between Microsoft
Entra ID and on-premises AD. Therefore, manual mapping with the second approach
isn't recommended.

The following sections describe the two mapping approaches.

Manual user remapping in Power BI

You can configure custom UPN rules in Power BI for Analysis Services data sources.
Custom rules help if your Power BI service sign-in name doesn't match your local
directory UPN. For example, if you sign in to Power BI with [email protected] but your
local directory UPN is [email protected] , you can configure a mapping rule to pass
[email protected] to Analysis Services.

） Important

The mapping works for the specific data source that's being configured. It's not a
global setting. If you have multiple Analysis Services data sources, you have to map
the users for each data source.

To do manual UPN mapping, follow these steps:

1. Under the Power BI gear icon, select Manage gateways and connections.

2. Select the data source, and then select Settings from the top menu.

3. On the Settings screen, in the Map user names box, make sure
EffectiveUserName is selected and then select Add new rule.

4. Under Map user names, for each user name to map, enter values for Original
name and New name, and then select Add new rule. The Replace value is the
sign-in address for Power BI, and the With value is the value to replace it with. The
replacement passes to the EffectiveUserName property for the Analysis Services
connection.
For example:
 ７ Note

Be sure not to change users that you don't intend to change. For example, if
you replace the Original name of contoso.com with a New name of
@contoso.local , all user sign-ins that contain @contoso.com are replaced with

@contoso.local . Also, if you replace an Original name of [email protected]

with a New name of [email protected] , a sign-in of [email protected]

is sent as [email protected] .

You can select an item in the list and reorder it by dragging and dropping, or
delete an entry by selecting the garbage can icon.

Use a wildcard
You can use a * wildcard for your Replace (original name) string. You can only use the
wildcard on its own and not with any other string part. Use a wildcard if you want to
replace all users with a single value to pass to the data source. This approach is useful
when you want all users in an organization to use the same user in your local
environment.

Test the mapping rule

To validate the name replacement, enter a value for Original name, and select Test rule.

７ Note

The saved rules work immediately in the browser. It take a few minutes before the
Power BI service starts to use the saved rules.

Active Directory lookup mapping

This section describes how to do an on-premises Active Directory property lookup to
remap Microsoft Entra UPNs to AD users. First, review how this remapping works.

Each query by a Power BI Microsoft Entra user to an on-premises SSAS server passes
along a UPN string such as [email protected] .

Lookup mapping in an on-premises data gateway with configurable custom user

mapping follows these steps:

1. Find the Active Directory to search. You can use automatic or configurable.
2. Look up the attribute of the Active Directory user, such as Email, from the Power BI
service. The attribute is based on an incoming UPN string like
[email protected] .
3. If the Active Directory lookup fails, it attempts to pass along the UPN to SSAS as
the EffectiveUserName .
4. If the Active Directory lookup succeeds, it retrieves the UserPrincipalName of that
Active Directory user.
 5. The mapping passes the UserPrincipalName email, such as [email protected]
prem.contoso , to SSAS as the EffectiveUserName .

７ Note

Any manual UPN user mappings defined in the Power BI data source gateway
configuration are applied before sending the UPN string to the on-premises data
gateway.

For the Active Directory lookup to work properly at runtime, you must change the on-
premises data gateway service to run with a domain account instead of a local service
account.

1. Make sure to download and install the latest gateway.

2. In the On-premises data gateway app on your machine, go to Service settings >
Change service account. Make sure you have the recovery key for the gateway,
because you need to restore it on the same machine unless you want to create a
new gateway. You must restart the gateway service for the change to take effect.

3. Go to the gateway's installation folder, C:\Program Files\On-premises data gateway,

as an administrator to ensure that you have write permissions. Open the
Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll.config file.

4. Edit the ADUserNameLookupProperty and the ADUserNameReplacementProperty values

according to the AD attribute configurations for your AD users. The values in the
following image are examples. These configurations are case sensitive, so make
sure they match the values in AD.

If the file provides no value for the ADServerPath configuration, the gateway uses
the default global catalog. You can specify multiple values for the ADServerPath .
The values must be separated by semicolons, as in the following example:

XML

<setting name="ADServerPath" serializeAs="String">

<value> GC://serverpath1; GC://serverpath2;GC://serverpath3</value>
 </setting>

The gateway parses the values for ADServerPath from left to right until it finds a
match. If the gateway doesn't find a match, it uses the original UPN. Make sure the
account that runs the gateway service, PBIEgwService, has query permissions to all
AD servers that you specify in ADServerPath .

The gateway supports two types of ADServerPath :

For WinNT: <value="WinNT://usa.domain.corp.contoso.com,computer"/>

For global catalog (GC): <value> GC://USA.domain.com </value>

5. Restart the on-premises data gateway service for the configuration change to take
effect.

Authentication to a live Analysis Services data

source
Each time a user interacts with Analysis Services, the effective user name is passed to the
gateway and then to the on-premises Analysis Services server. The UPN, which is
typically the email address you use to sign in to the cloud, is passed to Analysis Services
as the effective user in the EffectiveUserName connection property.

When the dataset is in Import Mode, the gateway will send the EffectiveUserName of
the UPN of the dataset owner. This means that the UPN of the dataset owner will be
passed to Analysis Services as the effective user in the EffectiveUserName connection
property.

This email address should match a defined UPN within the local Active Directory
domain. The UPN is a property of an AD account. A Windows account must be present
in an Analysis Services role to have access to the server. If no match is found in Active
Directory, the sign-in won't be successful.

Role-based and row-level security

Analysis Services can also provide filtering based on the Active Directory account. The
filtering can use role-based security or row-level security. A user's ability to query and
view model data depends on the roles that their Windows user account belongs to, and
on dynamic row-level security if it's configured.
 Role-based security. Models provide security based on user roles. You can define
roles for a particular model project during authoring in SQL Server Data Tools
Business Intelligence tools. After a model is deployed, you can define roles by
using SQL Server Management Studio. Roles contain members assigned by
Windows user name or by Windows group.

Roles define the permissions users have to query or take actions on the model.
Most users belong to a role with read permissions. Other roles give administrators
permissions to process items, manage database functions, and manage other roles.

Row-level security. Models can provide dynamic row-level security. Any defined
row-level security is specific to Analysis Services. For role-based security, every user
must have at least one role, but no tabular model requires dynamic row-level
security.

At a high level, dynamic security defines a user's read access to data in particular
rows in particular tables. Similar to roles, dynamic row-level security relies on a
user's Windows user name.

Implementing role and dynamic row-level security in models is beyond the scope of this
article. For more information, see Roles in tabular models and Security roles (Analysis
Services - Multidimensional data). For the most in-depth understanding of tabular
model security, download the Securing the tabular BI semantic model whitepaper.

Microsoft Entra authentication

Microsoft cloud services use Microsoft Entra ID to authenticate users. Microsoft Entra ID
is the tenant that contains user names and security groups. Typically, the email address a
user signs in with is the same as the UPN of the account.

Roles in the local Active Directory instance

For Analysis Services to determine if a user belongs to a role with permissions to read
data, the server needs to convert the effective user name passed from Microsoft Entra ID
to the gateway and on to the Analysis Services server. The Analysis Services server
passes the effective user name to a Windows Active Directory domain controller (DC).
The Active Directory DC then validates that the effective user name is a valid UPN on a
local account. The DC returns the user's Windows user name back to the Analysis
Services server.

You can't use EffectiveUserName on a non-domain joined Analysis Services server. The
Analysis Services server must be joined to a domain to avoid sign-in errors.
Identify your UPN
You might not know what your UPN is, and you might not be a domain administrator.
You can use the following command from your workstation to find out the UPN for your
account:

Windows Command Prompt

whoami /upn

The result looks similar to an email address, but is the UPN that's on your domain
account. If you use an Analysis Services data source for live connections, and this UPN
doesn't match the email address you use to sign in to Power BI, you might need to map
your user name.

Synchronize an on-premises AD with Microsoft Entra ID

If you plan to use Analysis Services live connections, your local AD accounts must match
Microsoft Entra ID. The UPN must match between the accounts.

Cloud services only use accounts within Microsoft Entra ID. If you add an account in
your local AD instance that doesn't exist in Microsoft Entra ID, you can't use the account.
There are several ways you can match your local AD accounts with Microsoft Entra ID:

Add accounts manually to Microsoft Entra ID.

Create an account on the Azure portal, or within the Microsoft 365 admin center,
with an account name that matches the UPN of the local AD account.

Use Microsoft Entra Connect Sync to synchronize local accounts to your Microsoft
Entra tenant.

Microsoft Entra Connect ensures that the UPN matches between Microsoft Entra ID
and your local AD instance. The Microsoft Entra Connect tool provides options for
directory synchronization and setting up authentication. Options include password
hash sync, pass-through authentication, and federation. If you're not an admin or a
local domain administrator, contact your IT admin to help with configuration.

７ Note

Synchronizing accounts with Microsoft Entra Connect Sync creates new

accounts within your Microsoft Entra tenant.
Use the data source
After you add the SSAS data source, it's available to use with either live connections or
through scheduled refresh.

７ Note

The server and database name must match between Power BI Desktop and the
data source within the on-premises data gateway.

The link between your dataset and the data source within the gateway is based on your
server name and database name. These names must match. For example, if you supply
an IP address for the server name within Power BI Desktop, you must use the IP address
for the data source within the gateway configuration. If you use SERVER\INSTANCE in
Power BI Desktop, you also must use SERVER\INSTANCE within the data source
configured for the gateway. This requirement holds for both live connections and
scheduled refresh.

Use the data source with live connections

You can use a live connection against tabular or multidimensional instances. You select a
live connection in Power BI Desktop when you first connect to the data. Make sure that
the server and database name matches between Power BI Desktop and the configured
data source for the gateway. Also, to be able to publish live connection datasets, your
users must appear under Users in the data source list.

After you publish reports, either from Power BI Desktop or by getting data in the Power
BI service, your data connection should start to work. It might take several minutes after
you create the data source in the gateway before you can use the connection.

Use the data source with scheduled refresh

If you're listed in the Users tab of the data source configured within the gateway, and
the server and database name match, you see the gateway as an option to use with
scheduled refresh.
Limitations of Analysis Services live connections
Cell level formatting and translation features aren't supported.

Actions and named sets aren't exposed to Power BI. You can still connect to
multidimensional cubes that contain actions or named sets to create visuals and
reports.

SKU requirements

ﾉ Expand table

Server version Required SKU

2012 SP1 CU4 or later Business Intelligence and Enterprise SKU

2014 Business Intelligence and Enterprise SKU

2016 Standard SKU or higher

Next steps
Troubleshoot the on-premises data gateway
Troubleshoot gateways - Power BI

More questions? Try the Power BI Community .

Manage your data source - SAP HANA
Article • 02/22/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

After you install the on-premises data gateway, you need to add data sources that can
be used with the gateway. This article looks at how to work with gateways and SAP
HANA data sources that are used either for scheduled refresh or for DirectQuery.

Add a data source

For more information about how to add a data source, see Add a data source. Under
Connection type, select SAP HANA.
After you select the SAP HANA data source type, fill in the Server, Username, and
Password information for the data source.

７ Note

All queries to the data source run using these credentials. To learn more about how
credentials are stored, see Store encrypted credentials in the cloud.
After you fill in everything, select Create. You can now use this data source for scheduled
refresh or DirectQuery against an SAP HANA server that is on-premises. You see Created
New data source if it succeeded.

Advanced settings
Optionally, you can configure the privacy level for your data source. This setting controls
how data can be combined. It's only used for scheduled refresh. The privacy-level
setting doesn't apply to DirectQuery. To learn more about privacy levels for your data
source, see Set privacy levels (Power Query) .
Use the data source
After you create the data source, it's available to use with either DirectQuery
connections or through scheduled refresh.

７ Note

The server and database names must match between Power BI Desktop and the
data source within the on-premises data gateway.

The link between your dataset and the data source within the gateway is based on your
server name and database name. These names must match. For example, if you supply
an IP address for the server name within Power BI Desktop, you must use the IP address
for the data source within the gateway configuration. If you use SERVER\INSTANCE in
Power BI Desktop, you also must use it within the data source configured for the
gateway.

This requirement is the case for both DirectQuery and scheduled refresh.

Use the data source with DirectQuery connections

Make sure that the server and database names match between Power BI Desktop and
the configured data source for the gateway. You also need to make sure your user is
listed in the Users tab of the data source to publish DirectQuery datasets. The selection
for DirectQuery occurs within Power BI Desktop when you first import data. For more
information about how to use DirectQuery, see Use DirectQuery in Power BI Desktop.

After you publish, either from Power BI Desktop or Get Data, your reports should start
to work. It might take several minutes after you create the data source within the
gateway for the connection to be usable.
Use the data source with scheduled refresh
If you're listed in the Users tab of the data source configured within the gateway and
the server name and database name match, you see the gateway as an option to use
with scheduled refresh.

Next steps
Troubleshoot the on-premises data gateway
Troubleshoot gateways - Power BI

More questions? Try asking the Power BI Community .

Manage a SQL Server data source
Article • 03/21/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

After you install an on-premises data gateway, you can add data sources to use with the
gateway. This article describes how to add a SQL Server data source to an on-premises
data gateway to use for scheduled refresh or DirectQuery.

Add a data source

Follow these instructions to add a SQL Server data source to your on-premises data
gateway.

７ Note

When you use DirectQuery, the gateway supports only SQL Server 2012 SP1 and
later.

1. On the New connection screen, select On-premises. Enter the Gateway cluster
name and new Connection name, and under Connection type, select SQL Server.
2. Fill in the Server and Database information for the data source.

3. Under Authentication Method, choose either Windows or Basic. Choose Basic if

you plan to use SQL authentication instead of Windows authentication. Then enter
the credentials to use for this data source.
 All queries to the data source run using these credentials unless you configure and
enable Kerberos single sign-on (SSO) for the data source. With SSO, datasets use
the current Power BI user's SSO credentials to execute the queries.

For more information about storing and using credentials, see:

Store encrypted credentials in the cloud

Use Kerberos for single sign-on (SSO) from Power BI to on-premises data
sources.

4. Configure the Privacy level for your data source. This setting controls how data can
be combined for scheduled refresh only. The privacy level setting doesn't apply to
DirectQuery. To learn more about privacy levels for your data source, see Privacy
levels (Power Query) .

5. Select Create.
You see a success message if the creation succeeds. You can now use this data source
for scheduled refresh or DirectQuery against an on-premises SQL Server.

For more information about how to add a data source, see Add a data source.

Use the data source

After you create the data source, it's available to use with either DirectQuery
connections or through scheduled refresh.

Server and database names must match

The link between your dataset and the data source in the gateway is based on your
server name and database name. These names must match exactly.
For example, if you supply an IP address for the server name in Power BI Desktop, you
must use the IP address for the data source in the gateway configuration. If you use
SERVER\INSTANCE in Power BI Desktop, you must use SERVER\INSTANCE in the data
source you configure for the gateway. This requirement holds for both DirectQuery and
scheduled refresh.

Use the data source with DirectQuery connections

Make sure that the server and database names match between Power BI Desktop and
the configured data source for the gateway. Also, to be able to publish DirectQuery
datasets, your users must appear under Users in the data source list.

You select the DirectQuery connection method in Power BI Desktop when you first
connect to data. For more information about how to use DirectQuery, see Use
DirectQuery in Power BI Desktop.

After you publish reports, either from Power BI Desktop or by getting data in Power BI
service, your SQL Server on-premises data connection should work. It might take several
minutes after you create the data source in the gateway to be able to use the
connection.

Use the data source with scheduled refresh

If you're listed in the Users column of the data source configured within the gateway,
and the server name and database name match, you see the gateway as an option to
use with scheduled refresh.

Next steps
Connect to on-premises data in SQL Server
 Troubleshoot the on-premises data gateway
Troubleshoot gateways - Power BI
Use Kerberos for single sign-on (SSO) from Power BI to on-premises data sources

More questions? Try asking the Power BI Community .

Manage your data source - Oracle
Article • 09/27/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

After you install the on-premises data gateway, you can add data sources to use with
the gateway. This article looks at how to work with the on-premises gateway and Oracle
data sources either for scheduled refresh or for DirectQuery.

Connect to an Oracle database

To connect to an Oracle database with the on-premises data gateway, download and
install the 64-bit Oracle Client for Microsoft Tools (OCMT) on the computer running
the gateway.

Supported Oracle versions are:

Oracle Database Server 12c (12.1.0.2) and later

Oracle Autonomous Database - all versions

After you install and configure OCMT properly, you can use Power BI Desktop or
another test client to verify the correct installation and configuration on the gateway.

Add a data source

1. On the New connection screen for your on-premises data gateway, select Oracle
for Connection type.
2. In Server, enter the name for the data source, such as your Oracle net service
name (for example, myADB_high) or Easy Connect Plus connection string.

3. Under Authentication method, choose either Windows or Basic. Choose Basic if

you plan to log in as an Oracle database user. Then enter the credentials to use for
this data source. Choose Windows when using Windows operating system
authentication and with both the Oracle client and server running on Windows.

７ Note

All queries to the data source run with these credentials. To learn more about
credential storage, see Store encrypted credentials in the cloud.

4. Configure the Privacy level for your data source. This setting controls how data can
combine for scheduled refresh. The privacy-level setting doesn't apply to
DirectQuery. To learn more about privacy levels for your data source, see Privacy
levels (Power Query) .

5. Select Create.
 If the creation succeeds, you see Created <Data source name>. You can now use
this data source for scheduled refresh or DirectQuery with the Oracle database
server.

Use the data source

After you create the data source, it's available to use with either DirectQuery or
scheduled refresh.

） Important

The server and database names must match between Power BI Desktop and the
data source within the on-premises data gateway.

The link between your dataset and the data source within the gateway is based on your
server name and database name. These names must match exactly. For example, if you
supply an IP address for the server name within Power BI Desktop, you must use the IP
address for the data source within the gateway configuration. This name also has to
match a net service name or alias that the tnsnames.ora file defines. This requirement is
the case for both DirectQuery and scheduled refresh.

Use the data source with DirectQuery connections

Make sure that the server and database names match between Power BI Desktop and
the configured data source for the gateway. Also, to be able to publish DirectQuery
datasets, your users must appear under Users in the data source listing.

After you publish reports, either from Power BI Desktop or by getting data in Power BI
service, your database connection should work. It might take several minutes after you
create the data source in the gateway to be able to use the connection.

Use the data source with scheduled refresh

If you're in the Users list of a data source you configure within the gateway, and the
server and database names match, you see the gateway as an option to use with
scheduled refresh.
Troubleshooting
You might get one of the following Oracle errors when the naming syntax is either
incorrect or improperly configured:

ORA-12154: TNS:could not resolve the connect identifier specified.

ORA-12514: TNS:listener does not currently know of service requested in

connect descriptor.
ORA-12541: TNS:no listener.

ORA-12170: TNS:connect timeout occurred.

ORA-12504: TNS:listener was not given the SERVICE_NAME in CONNECT_DATA.

These errors might occur if the Oracle tnsnames.ora database connect descriptor is
misconfigured, the net service name provided is misspelled, or the Oracle database
listener is not running or not reachable, such as a firewall blocking the listener or
database port. Be sure you are meeting the minimum installation prerequisites.

Visit the Oracle Database Error Help Portal to review common causes and resolutions
for the specific Oracle error you encounter. Enter your Oracle error in the portal search
bar.

To diagnose connectivity issues between the data source server and the gateway
machine, install a client like Power BI Desktop on the gateway machine. You can use the
client to check connectivity to the data source server.

For more gateway troubleshooting information, see Troubleshoot the on-premises data
gateway.

Next steps
Troubleshoot gateways - Power BI
Power BI Premium

More questions? Try asking the Power BI Community .

Manage your data source - import and
scheduled refresh
Article • 03/07/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

After you install the on-premises data gateway, you need to add data sources that can
be used with the gateway. This article looks at how to work with gateways and data
sources that are used for scheduled refresh as opposed to DirectQuery or live
connections.

Add a data source

Select a data source type. All of the data source types listed can be used for scheduled
refresh with the on-premises data gateway. Analysis Services, SQL Server, and SAP
HANA can be used for scheduled refresh, DirectQuery, or live connections. For more
information about how to add a data source, see Add a data source.
Then fill in the information for the data source, which includes the source information
and credentials that are used to access the data source.

７ Note

All queries to the data source run by using these credentials. To learn more about
how credentials are stored, see Store encrypted credentials in the cloud.
For a list of data source types that can be used with scheduled refresh, see List of
available data source types.

After you fill in everything, select Create. If the action succeeds, you see Created New
data source. You can now use this data source for scheduled refresh with your on-
premises data.
Advanced settings
Optionally, you can configure the privacy level for your data source. This setting controls
how data can be combined. It's only used for scheduled refresh. To learn more about
privacy levels for your data source, see Privacy levels (Power Query) .

Use the data source for scheduled refresh

After you create the data source, it's available to use with either DirectQuery
connections or through scheduled refresh.

７ Note

The server and database names must match between Power BI Desktop and the
data source within the on-premises data gateway.

The link between your dataset and the data source within the gateway is based on your
server name and database name. These names must match. For example, if you supply
an IP address for the server name within Power BI Desktop, you must use the IP address
for the data source within the gateway configuration. If you use SERVER\INSTANCE in
Power BI Desktop, you also must use it within the data source configured for the
gateway.

If you're listed in the Users tab of the data source configured within the gateway and
the server name and database name match, you see the gateway as an option to use
with scheduled refresh.
 ） Important

Upon republish, the data set owner must associate the dataset to a gateway and
corresponding data source again. The previous association is not maintained after
republishing.

２ Warning

If your dataset contains multiple data sources, each data source must be added
within the gateway. If one or more data sources aren't added to the gateway, you
don't see the gateway as available for scheduled refresh.

Next steps
Troubleshooting the on-premises data gateway
Troubleshoot gateways - Power BI

More questions? Try the Power BI Community .

Merge or append on-premises and
cloud data sources
Article • 05/28/2024

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

You can use the on-premises data gateway to merge or append on-premises and cloud
data sources in the same query. This solution is helpful when you want to combine data
from multiple sources without having to use separate queries.

７ Note

This article applies only to datasets that have cloud and on-premises data sources
merged or appended in a single query. For datasets that include separate queries,
for instance, one that connects to an on-premises data source and the other to a
cloud data source, the gateway doesn't execute the query for the cloud data
source.

Prerequisites
A gateway installed on a local computer.
A Power BI Desktop file with queries that combine on-premises and cloud data
sources.

７ Note

To access any cloud data sources, you must ensure that the gateway has access to
those data sources.

1. In the upper-right corner of the Power BI service, select the gear icon then
Manage connections and gateways.
2. Select the gateway you want to configure, and select Settings from the top ribbon

3. Under Settings, select Allow user's cloud data sources to refresh through this
gateway cluster, then select Save.
 4. To add any on-premises data sources used in your queries, select Connections,
then select New to create a connection. You don't need to add the cloud data
sources here.

5. Select your gateway for Gateway cluster name. Name the connection and specify
the type of connection and other required information. Then select Create.

6. Upload to the Power BI service your Power BI Desktop file with the queries that
combine on-premises and cloud data sources.

With the cloud credentials set, you can now refresh the dataset by using the Refresh
now option. Or, you can schedule it to refresh periodically.

Related content
To learn more about data refresh for gateways, see: Use the data source for scheduled
refresh.

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Overview of single sign-on for on-
premises data gateways in Power BI
Article • 01/23/2024

You can get seamless single sign-on (SSO) connectivity, enabling Power BI reports and
dashboards to update in real time by configuring your on-premises data gateway. You
have the option of configuring your gateway with the following SSO options:

Active Directory (AD) SSO, which includes:

Kerberos constrained delegation.
Security Assertion Markup Language (SAML).
Microsoft Entra SSO.

７ Note

SSO is only supported by Power BI datasets and not by Power BI dataflows.

Supported data sources for SSO

AD SSO is usually configured for on-premises data sources that are secured within your
on-premises network. Microsoft Entra SSO is configured for data sources that support
Microsoft Entra authentication, typically cloud data sources, secured behind an Azure
Virtual Network.

While the on-premises data gateway supports SSO by using DirectQuery or Refresh for
the AD-based SSO options, only DirectQuery is supported for Microsoft Entra SSO.

Power BI supports the following data sources:

Amazon Redshift (Microsoft Entra ID)

Azure Databricks
Azure Data Explorer (Microsoft Entra ID)
Azure SQL (Microsoft Entra ID)
Azure Synapse Analytics (Microsoft Entra ID)
Denodo (Kerberos)
Hive LLAP (Kerberos)
Impala (Kerberos)
Oracle (Kerberos)
SAP BW Application Server (Kerberos)
 SAP BW Message Server (Kerberos)
SAP HANA (Kerberos and SAML)
Snowflake (Microsoft Entra ID)
Spark (Kerberos)
SQL Server (Kerberos)
Teradata (Kerberos)
Tibco Data Virtualization (Kerberos)

７ Note

SQL Server Analysis Services also supports SSO, but does so using Live
connections, rather than using Kerberos or SAML. Power BI doesn't support SSO for
M-extensions.

Interact with reports that rely on SSO

When a user interacts with a DirectQuery report in the Power BI service, each cross-filter,
slice, sort, and report editing operation can result in queries that execute live against the
underlying data source. When you configure SSO for the data source, queries execute
under the identity of the user that interacts with Power BI. That is, they run through the
web experience or Power BI mobile apps. Therefore, each user sees precisely the data
for which they have permissions in the underlying data source.

You can also configure a report that is set up for refresh in the Power BI service to use
SSO. When you configure SSO for this data source, queries execute under the identity of
the dataset owner within Power BI. Therefore, the refresh happens based on the dataset
owner's permissions on the underlying data source. Refresh using SSO is currently
enabled only for data sources using Kerberos constrained delegation.

Related content
Now that you understand the basics of SSO through the gateway, read detailed
information about setting up SSO here:

Active Directory (AD) SSO

Microsoft Entra SSO
Active Directory (AD) SSO
Article • 01/23/2024

The on-premises data gateway supports Active Directory (AD) SSO for connecting to
your on-premises data sources that have Active Directory configured. AD SSO includes
both Kerberos constrained delegation and Security Assertion Markup Language (SAML).
For more information on SSO and the list of data sources supported for AD SSO, see
Overview of single sign-on (SSO) for on-premises data gateways in Power BI.

Query steps when running Active Directory

SSO
A query that runs with SSO consists of three steps, as shown in the following diagram.

Here are additional details about each step:

1. For each query, the Power BI service includes the user principal name (UPN), which
is the fully qualified username of the user currently signed in to the Power BI
service, when it sends a query request to the configured gateway.

2. The gateway must map the Microsoft Entra UPN to a local Active Directory identity:

a. If Microsoft Entra DirSync (also known as Microsoft Entra Connect) is configured,

then the mapping works automatically in the gateway.
 b. Otherwise, the gateway can look up and map the Microsoft Entra UPN to a local
AD user by performing a lookup against the local Active Directory domain.

3. The gateway service process impersonates the mapped local user, opens the
connection to the underlying database, and then sends the query. You don't need
to install the gateway on the same machine as the database.

Related content
Now that you understand the basics of enabling SSO through the gateway, read more
detailed information about Kerberos and SAML:

Single sign-on (SSO) - Kerberos

Single sign-on (SSO) - SAML
Overview of single sign-on (SSO) for on-premises data gateways in Power BI
Configure Kerberos-based SSO from
Power BI service to on-premises data
sources
Article • 01/23/2024

Enabling SSO makes it easy for Power BI reports and dashboards to refresh data from
on-premises sources while respecting user-level permissions configured on those
sources. Use Kerberos constrained delegation to enable seamless SSO connectivity.

This article describes the steps you need to take to configure Kerberos-based SSO from
Power BI service to on-premises data sources.

Prerequisites
Several items must be configured for Kerberos constrained delegation to work properly,
including *Service Principal Names (SPN) and delegation settings on service accounts.

７ Note

Using DNS aliasing with SSO is not supported.

Configuration outline
The steps required for configuring gateway single sign-on are outlined below.

1. Complete all the steps in Section 1: Basic configuration.

2. Depending on your Active Directory environment and the data sources used, you
may need to complete some or all of the configuration described in Section 2:
Environment-specific configuration.

Possible scenarios that may require additional configuration are listed below:

ﾉ Expand table

Scenario Go to

Your Active Directory environment is security hardened. Add gateway service

account to Windows
 Scenario Go to

Authorization and Access

Group

The gateway service account and the user accounts that the Add gateway service
gateway will impersonate are in separate domains or forests. account to Windows
Authorization and Access
Group

You don't have Microsoft Entra Connect with user account Set user-mapping
synchronization configured and the UPN used in the Power configuration parameters on
BI for users does not match the UPN in your local Active the gateway machine
Directory environment.

You plan to use an SAP HANA data source with SSO. Complete data source-
specific configuration steps

You plan to use an SAP BW data source with SSO. Complete data source-
specific configuration steps

You plan to use a Teradata data source with SSO. Complete data source-
specific configuration steps

3. Validate your configuration as described in Section 3: Validate configuration to

ensure that SSO is set up correctly.

Section 1: Basic configuration

Step 1: Install and configure the Microsoft on-premises

data gateway
The on-premises data gateway supports an in-place upgrade, and settings takeover of
existing gateways.

Step 2: Obtain domain admin rights to configure SPNs

(SetSPN) and Kerberos constrained delegation settings
To configure SPNs and Kerberos delegation settings, a domain administrator should
avoid granting rights to someone that doesn't have domain admin rights. In the
following section, we cover the recommended configuration steps in more detail.

Step 3: Configure the Gateway service account

Option A below is the required configuration unless you have both Microsoft Entra
Connect configured and user accounts are synchronized. In that case, option B is
recommended.

Option A: Run the gateway Windows service as a domain account

with SPN

In a standard installation, the gateway runs as the machine-local service account, NT

Service\PBIEgwService.

To enable Kerberos constrained delegation, the gateway must run as a domain account,
unless your Microsoft Entra instance is already synchronized with your local Active
Directory instance (by using Microsoft Entra DirSync/Connect). To switch to a domain
account, see change the gateway service account.

Configure an SPN for the gateway service account

First, determine whether an SPN was already created for the domain account used as
the gateway service account:

1. As a domain administrator, launch the Active Directory Users and Computers

Microsoft Management Console (MMC) snap-in.

2. In the left pane, right-click the domain name, select Find, and then enter the
account name of the gateway service account.

3. In the search result, right-click the gateway service account and select Properties.

4. If the Delegation tab is visible on the Properties dialog, then an SPN was already
created and you can skip to Configure Kerberos constrained delegation.

5. If there isn't a Delegation tab on the Properties dialog box, you can manually
create an SPN on the account to enable it. Use the setspn tool that comes with
Windows (you need domain admin rights to create the SPN).

For example, suppose the gateway service account is Contoso\GatewaySvc and

the gateway service is running on the machine named MyGatewayMachine. To set
the SPN for the gateway service account, run the following command:

setspn -S gateway/MyGatewayMachine Contoso\GatewaySvc

 You can also set the SPN by using the Active Directory Users and Computers
MMC snap-in.

Option B: Configure computer for Microsoft Entra Connect

If Microsoft Entra Connect is configured and user accounts are synchronized, the
gateway service doesn't need to perform local Microsoft Entra lookups at runtime.
Instead, you can simply use the local service SID for the gateway service to complete all
required configuration in Microsoft Entra ID. The Kerberos constrained delegation
configuration steps outlined in this article are the same as the configuration steps
required in the Microsoft Entra context. They are applied to the gateway's computer
object (as identified by the local service SID) in Microsoft Entra ID instead of the domain
account. The local service SID for NT SERVICE/PBIEgwService is as follows:

S-1-5-80-1835761534-3291552707-3889884660-1303793167-3990676079

To create the SPN for this SID against the Power BI Gateway computer, you would need
to run the following command from an administrative command prompt (replace
<COMPUTERNAME> with the name of the Power BI Gateway computer):

SetSPN -s HTTP/S-1-5-80-1835761534-3291552707-3889884660-1303793167-3990676079

<COMPUTERNAME>

７ Note

Depending on your local security settings, you may need to add the gateway
service account, NT SERVICE\PBIEgwService, to the local Administrators group on
the gateway machine and then restart the gateway service in the gateway app.

Step 4: Configure Kerberos constrained delegation

You can configure delegation settings for either standard Kerberos constrained
delegation or resource-based Kerberos constrained delegation. For more information on
the differences between the two approaches to delegation, see Kerberos constrained
delegation overview.

The following service accounts are required:

Gateway service account: Service user representing the gateway in Active Directory,
with an SPN configured in Step 3.
 Data Source service account: Service user representing the data source in Active
Directory, with an SPN mapped to the data source.

７ Note

The gateway and data source service accounts must be separate. The same service
account cannot be used to represent both the gateway and data source.

Depending on which approach you want to use, proceed to one of the following
sections. Don't complete both sections:

Option A: Standard Kerberos constrained delegation. This is the default

recommendation for most environments.
Option B: Resource-based Kerberos constrained delegation. This is required if your
data source belongs to a different domain than your gateway.

Option A: Standard Kerberos constrained delegation

We'll now set the delegation settings for the gateway service account. There are multiple
tools you can use to perform these steps. Here, we'll use the Active Directory Users and
Computers MMC snap-in to administer and publish information in the directory. It's
available on domain controllers by default; on other machines, you can enable it
through Windows feature configuration.

We need to configure Kerberos constrained delegation with protocol transition. With

constrained delegation, you must be explicit about which services you allow the
gateway to present delegated credentials to. For example, only SQL Server or your SAP
HANA server accepts delegation calls from the gateway service account.

This section assumes you have already configured SPNs for your underlying data
sources (such as SQL Server, SAP HANA, SAP BW, Teradata, or Spark). To learn how to
configure those data source server SPNs, refer to the technical documentation for the
respective database server and see the section What SPN does your app require? in the
My Kerberos Checklist blog post.

In the following steps, we assume an on-premises environment with two machines in

the same domain: a gateway machine and a database server running SQL Server that
has already been configured for Kerberos-based SSO. The steps can be adopted for one
of the other supported data sources, so long as the data source has already been
configured for Kerberos-based single sign-on. For this example, we'll use the following
settings:
 Active Directory Domain (Netbios): Contoso
Gateway machine name: MyGatewayMachine
Gateway service account: Contoso\GatewaySvc
SQL Server data source machine name: TestSQLServer
SQL Server data source service account: Contoso\SQLService

Here's how to configure the delegation settings:

1. With domain administrator rights, open the Active Directory Users and
Computers MMC snap-in.

2. Right-click the gateway service account (Contoso\GatewaySvc), and select

Properties.

3. Select the Delegation tab.

4. Select Trust this computer for delegation to specified services only > Use any
authentication protocol.

5. Under Services to which this account can present delegated credentials, select
Add.

6. In the new dialog box, select Users or Computers.

7. Enter the service account for the data source, and then select OK.

For example, a SQL Server data source can have a service account like
Contoso\SQLService. An appropriate SPN for the data source should have already
been set on this account.

8. Select the SPN that you created for the database server.

In our example, the SPN begins with MSSQLSvc. If you added both the FQDN and
the NetBIOS SPN for your database service, select both. You might see only one.

9. Select OK.

You should now see the SPN in the list of services to which the gateway service
account can present delegated credentials.
 10. To continue the setup process, proceed to Grant the gateway service account local
policy rights on the gateway machine.

Option B: Resource-based Kerberos constrained delegation

You use resource-based Kerberos constrained delegation to enable single sign-on

connectivity for Windows Server 2012 and later versions. This type of delegation permits
front-end and back-end services to be in different domains. For it to work, the back-end
service domain needs to trust the front-end service domain.
In the following steps, we assume an on-premises environment with two machines in
different domains: a gateway machine and a database server running SQL Server that
has already been configured for Kerberos-based SSO. These steps can be adopted for
one of the other supported data sources, so long as the data source has already been
configured for Kerberos-based single sign-on. For this example, we'll use the following
settings:

Active Directory frontend Domain (Netbios): ContosoFrontEnd

Active Directory backend Domain (Netbios): ContosoBackEnd
Gateway machine name: MyGatewayMachine
Gateway service account: ContosoFrontEnd\GatewaySvc
SQL Server data source machine name: TestSQLServer
SQL Server data source service account: ContosoBackEnd\SQLService

Complete the following configuration steps:

1. Use the Active Directory Users and Computers MMC snap-in on the domain
controller for the ContosoFrontEnd domain and verify no delegation settings are
applied for the gateway service account.
2. Use Active Directory Users and Computers on the domain controller for the
ContosoBackEnd domain and verify no delegation settings are applied for the
back-end service account.

3. In the Attribute Editor tab of the account properties, verify that the msDS-
AllowedToActOnBehalfOfOtherIdentity attribute isn't set.
4. In Active Directory Users and Computers, create a group on the domain controller
for the ContosoBackEnd domain. Add the GatewaySvc gateway service account to
the ResourceDelGroup group.

To add users from a trusted domain, this group must have a scope of Domain
local.
 5. Open a command prompt and run the following commands in the domain
controller for the ContosoBackEnd domain to update the msDS-
AllowedToActOnBehalfOfOtherIdentity attribute of the back-end service account:

PowerShell

$c = Get-ADGroup ResourceDelGroup
Set-ADUser SQLService -PrincipalsAllowedToDelegateToAccount $c

6. In Active Directory Users and Computers, verify that the update is reflected in the
Attribute Editor tab in the properties for the back-end service account.

Step 5: Enable AES Encryption on Service Accounts

Apply the following settings to the gateway service account and every data source
service account that the gateway can delegate to:

７ Note

If there are existing enctypes defined on the service accounts(s), consult with your
Active Directory Administrator, because following the below steps will overwrite the
existing enctypes values and may break clients.

1. With domain administrator rights, open the Active Directory Users and
Computers MMC snap-in.

2. Right-click the gateway/data source service account and select Properties.

3. Select the Account tab.

4. Under Account Options, enable at least one (or both) of the following options.
Note that the same options need to be enabled for all service accounts.

This account supports Kerberos AES 128-bit encryption

This account supports Kerberos AES 256-bit encryption

７ Note

If you are unsure which encryption scheme to use, consult with your Active
Directory Administrator.

Step 6: Grant the gateway service account local policy

rights on the gateway machine
Finally, on the machine running the gateway service (MyGatewayMachine in our
example), grant the gateway service account the local policies Impersonate a client
after authentication and Act as part of the operating system (SeTcbPrivilege). Perform
this configuration with the Local Group Policy Editor (gpedit.msc).

1. On the gateway machine, run gpedit.msc.

2. Go to Local Computer Policy > Computer Configuration > Windows Settings >
Security Settings > Local Policies > User Rights Assignment.
 3. Under User Rights Assignment, from the list of policies, select Impersonate a
client after authentication.

4. Right-click the policy, open Properties, and then view the list of accounts.

The list must include the gateway service account (Contoso\GatewaySvc or

ContosoFrontEnd\GatewaySvc depending on the type of constrained delegation).

5. Under User Rights Assignment, select Act as part of the operating system
(SeTcbPrivilege) from the list of policies. Ensure that the gateway service account is
included in the list of accounts.

6. Restart the On-premises data gateway service process.

Step 7: Windows account can access gateway machine

SSO uses Windows Authentication, so make sure the Windows account can access the
gateway machine. If not sure, add NT-AUTHORITY\Authenticated Users (S-1-5-11) to the
local machine "Users" group.

Section 2: Environment-specific configuration

Add gateway service account to Windows Authorization

and Access Group
Complete this section if any of the following situations apply:

Your Active Directory environment is security hardened.

 When the gateway service account and the user accounts that the gateway will
impersonate are in separate domains or forests.

You can also add the gateway service account to Windows Authorization and Access
Group in situations where the domain / forest has not been hardened, but it isn't
required.

For more information, see Windows Authorization and Access Group.

To complete this configuration step, for each domain that contains Active Directory
users you want the gateway service account to be able to impersonate:

1. Sign in to a computer in the domain, and launch the Active Directory Users and
Computers MMC snap-in.
2. Locate the group Windows Authorization and Access Group, which is typically
found in the Builtin container.
3. Double click on the group, and click on the Members tab.
4. Click Add, and change the domain location to the domain that the gateway service
account resides in.
5. Type in the gateway service account name and click Check Names to verify that
the gateway service account is accessible.
6. Click OK.
7. Click Apply.
8. Restart the gateway service.

Set user-mapping configuration parameters on the

gateway machine
Complete this section if:

You don't have Microsoft Entra Connect with user account synchronization
configured AND
The UPN used in Power BI for users does not match the UPN in your local Active
Directory environment.

Each Active Directory user mapped in this way needs to have SSO permissions for your
data source.

1. Open the main gateway configuration file,

Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll . By default, this file is

stored at C:\Program Files\On-premises data gateway .

2. Set ADUserNameLookupProperty to an unused Active Directory attribute. We'll
use msDS-cloudExtensionAttribute1 in the steps that follow. This attribute is
available only in Windows Server 2012 and later.

3. Set ADUserNameReplacementProperty to SAMAccountName and then save the

configuration file.

７ Note

In multi-domain scenarios, you may need to set the

ADUserNameReplacementProperty to userPrincipalName to preserve the
domain information of the user.

4. From the Services tab of Task Manager, right-click the gateway service and select
Restart.

5. For each Power BI service user you want to enable Kerberos SSO for, set the msDS-
cloudExtensionAttribute1 property of a local Active Directory user (with SSO

permission to your data source) to the full username (UPN) of the Power BI service
user. For example, if you sign in to Power BI service as [email protected] and you
want to map this user to a local Active Directory user with SSO permissions, say,
[email protected], set this user's msDS-cloudExtensionAttribute1 attribute
to [email protected].

You can set the msDS-cloudExtensionAttribute1 property with the Active Directory
Users and Computers MMC snap-in:

a. As a domain administrator, launch Active Directory Users and Computers.

b. Right-click the domain name, select Find, and then enter the account name of
the local Active Directory user to map.

c. Select the Attribute Editor tab.

 Locate the msDS-cloudExtensionAttribute1 property, and double-click it. Set the
value to the full username (UPN) of the user you use to sign in to the Power BI
service.

d. Select OK.

e. Select Apply. Verify that the correct value has been set in the Value column.

Complete data source-specific configuration steps

For SAP HANA, SAP BW, and Teradata data sources, additional configuration is required
to use with gateway SSO:

Use Kerberos for single sign-on (SSO) to SAP HANA.

Use Kerberos single sign-on for SSO to SAP BW using CommonCryptoLib
(sapcrypto.dll).
Use Kerberos for single sign-on (SSO) to Teradata.

７ Note

Although other SNC libraries might also work for BW SSO, they aren't officially
supported by Microsoft.

Section 3: Validate configuration

Step 1: Configure data sources in Power BI

After you complete all the configuration steps, use the Manage Gateway page in Power
BI to configure the data source to use for SSO. If you have multiple gateways, ensure
that you select the gateway you've configured for Kerberos SSO. Then, under Settings
for the data source, ensure Use SSO via Kerberos for DirectQuery queries or Use SSO
via Kerberos for DirectQuery And Import queries is checked for DirectQuery based
Reports and Use SSO via Kerberos for DirectQuery And Import queries is checked for
Import based Reports.
The settings Use SSO via Kerberos for DirectQuery queries and Use SSO via Kerberos
for DirectQuery And Import queries give a different behavior for DirectQuery based
reports and Import based reports.

Use SSO via Kerberos for DirectQuery queries:

For DirectQuery based report, SSO credentials of the user are used.
For Import based report, SSO credentials are not used, but the credentials entered
in data source page are used.

Use SSO via Kerberos for DirectQuery And Import queries:

For DirectQuery based report, SSO credentials of the user are used.
For Import based report, the SSO credentials of the semantic model owner are
used, regardless of the user triggering the Import.

Step 2: Test single sign-on

Go to Test single sign-on (SSO) configuration to quickly validate that your configuration
is set correctly and troubleshoot common problems.

Step 3: Run a Power BI report

When you publish, select the gateway you've configured for SSO if you have multiple
gateways.

Related content
For more information about the on-premises data gateway and DirectQuery, see the
following resources:
What is an on-premises data gateway?
DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP BW
DirectQuery and SAP HANA
Use Kerberos for SSO to SAP HANA
Article • 03/17/2023

） Important

Because SAP no longer supports OpenSSL , Microsoft has also discontinued its
support. Your existing connections continue to work but you can no longer create
new connections. Use SAP Cryptographic Library (CommonCryptoLib), or sapcrypto,
instead.

This article describes how to configure your SAP HANA data source to enable single
sign-on (SSO) from the Power BI service.

７ Note

Before you attempt to refresh a SAP HANA-based report that uses Kerberos SSO,
complete the steps in both this article and Configure Kerberos SSO.

Enable SSO for SAP HANA

To enable SSO for SAP HANA, do the following steps:

1. Ensure the SAP HANA server is running the required minimum version, which
depends on your SAP HANA server platform level:

HANA 2 SPS 01 Rev 012.03

HANA 2 SPS 02 Rev 22
HANA 1 SP 12 Rev 122.13

2. On the gateway computer, install the latest SAP HANA ODBC driver. The minimum
version is HANA ODBC version 2.00.020.00 from August 2017.

3. Ensure that the SAP HANA server has been configured for Kerberos-based SSO.
For more information about setting up SSO for SAP HANA by using Kerberos, see
Single sign-on using Kerberos . Also see the links from that page, particularly SAP
Note 1837331 – HOWTO HANA DBSSO Kerberos/Active Directory.

We also recommend following these extra steps, which can yield a small performance
improvement:
 1. In the gateway installation directory, look for and open this configuration file:
Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll.config.

2. Look for the FullDomainResolutionEnabled property, and change its value to True .

XML

<setting name=" FullDomainResolutionEnabled " serializeAs="String">

<value>True</value>
</setting>

3. Run a Power BI report.

Troubleshoot
This section provides instructions for troubleshooting using Kerberos for single sign-on
(SSO) to SAP HANA in the Power BI service. By using these troubleshooting steps, you
can self-diagnose and correct many issues you might be facing.

To follow the steps in this section, you need to collect gateway logs.

TLS/SSL error (certificate)

This issue has multiple symptoms.

When you try to add a new data source, you might see an error like the following
message:

Output

Unable to connect: We encountered an error while trying to connect to.

Details: "We could not register this data source for any gateway
instances within this cluster.
Please find more details below about specific errors for each gateway
instance."

When you try to create or refresh a report, you might see the following error
message:
 

When you investigate the Mashup[date]*.log, you see the following error message:

Output

A connection was successfully established with the server,

but then an error occurred during the login process and
the certificate chain was issued by an authority that is not trusted.

Resolution

To resolve this TLS/SSL error, go to the data source connection and then, in the Validate
Server Certificate section, disable the setting, as shown in the following image:

After you've disabled this setting, the error message no longer appears.

Impersonation
Log entries for impersonation contain entries similar to:

Output

About to impersonate user DOMAIN\User (IsAuthenticated: True,

ImpersonationLevel: Impersonation).

The important element in this log entry is the information that's displayed after the
ImpersonationLevel: entry. Any value different from Impersonation reveals that

impersonation isn't occurring properly.

Resolution

You can set up ImpersonationLevel properly by following the instructions in Grant the
gateway service account local policy rights on the gateway.

After you've changed the configuration file, restart the gateway service for the change
to take effect.

Validation

Refresh or create the report, and then collect the gateway logs. Open the most recent
GatewayInfo file, and check the following string: About to impersonate user DOMAIN\User
(IsAuthenticated: True, ImpersonationLevel: Impersonation) . Make sure that the
ImpersonationLevel setting returns Impersonation .

Delegation
Delegation issues usually appear in the Power BI service as generic errors. To make sure
that the issue isn't a delegation issue, collect Wireshark traces and use Kerberos as a
filter. To learn more about Wireshark, and for information about Kerberos errors, see
Kerberos errors in network captures.

The following symptoms and troubleshooting steps can help remedy some common
issues.

SPN issues

If you see the following error: The import [table] matches no exports. Did you miss a
module reference?: while investigating the Mashup[date]*.log, then you're experiencing

service principal name (SPN) issues.

When you investigate further by using Wireshark traces, you reveal the error
KRB4KDC_ERR_S_PRINCIPAL_UNKOWN , which means that the SPN wasn't found or doesn't
exist. The following image shows an example:

Resolution

To resolve SPN issues such as this issue, you must add an SPN to a service account. For
more information, see the SAP documentation in Configure Kerberos for SAP HANA
database hosts .

In addition, follow the resolution instructions described in the next section.

No credentials issues

There might not be clear symptoms associated with this issue. When you investigate the
Mashup[date]*.log, you see the following error:

Output

29T20:21:34.6679184Z","Action":"RemoteDocumentEvaluator/RemoteEvaluation/Han
dleException","HostProcessId":"1396","identity":"DirectQueryPool","Exception
":"Exception:\r\nExceptionType:
Microsoft.Mashup.Engine1.Runtime.ValueException, Microsoft.MashupEngine,
Version=1.0.0.0, Culture=neutral,
PublicKeyToken=31bf3856ad364e35\r\nMessage:

When you investigate the same file further, the following (unhelpful) error appears:

Output
 No credentials are available in the security package

Capturing Wireshark traces reveals the following error: KRB5KDC_ERR_BADOPTION .

Usually, these errors mean that the SPN hdb/hana2-s4-sso2.westus2.cloudapp.azure.com

file could be found but isn't in the Services to which this account can present
delegated credentials list on the Delegation pane in the Gateway service account.

Resolution

To resolve the No credentials issue, follow the steps described in Configure Kerberos
constrained delegation. When completed properly, the delegation tab at the gateway
service account reflects the HansaWorld Database (HDB) file and fully qualified domain
name (FQDN) in the list of Services to which this account can present delegated
credentials.

Validation
Following the preceding steps should resolve the issue. If you still experience Kerberos
issues, you might have a misconfiguration in the Power BI gateway or in the HANA
server itself.

Credentials errors
If you experience credentials errors, errors in the logs or traces expose errors that
describe Credentials are invalid or similar errors. These errors might manifest
differently on the data source side of the connection, such as SAP HANA. The following
image shows an example error:
 

Symptom 1
In HANA authentication traces, you might see entries similar to the following message:

Output

[Authentication|manager.cpp:166] Kerberos: Using Service Principal

Name [email protected]@CONTOSO.COM with name type:
GSS_KRB5_NT_PRINCIPAL_NAME
[Authentication|methodgssinitiator.cpp:367] Got principal name:
[email protected]@CONTOSO.COM

Resolution
Follow the instructions described in Set user-mapping configuration parameters on the
gateway machine, even if you've already configured the Azure AD Connect service.

Validation
After you've completed the validation, you can successfully load the report in the Power
BI service.

Symptom 2
In HANA authentication traces, you might see entries similar to the following entry:

Output

Authentication ManagerAcceptor.cpp(00233) : Extending list of expected

external names by [email protected] (method: GSS) Authentication
AuthenticationInfo.cpp(00168) : ENTER getAuthenticationInfo
([email protected]) Authentication
AuthenticationInfo.cpp(00237) :
Found no user with expected external name!

Resolution
Check the Kerberos external ID under HANA User to determine whether the IDs match
properly.

Validation

After you've resolved the issue, you can create or refresh reports in the Power BI service.

Next steps
For more information about the on-premises data gateway and DirectQuery, see the
following resources:

What is an on-premises data gateway?

DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP Business Warehouse (BW)
DirectQuery and SAP HANA
Use Kerberos single sign-on for SSO to
SAP BW using CommonCryptoLib
(sapcrypto.dll)
Article • 03/29/2022

This article describes how to configure your SAP BW data source to enable SSO from the
Power BI service by using CommonCryptoLib (sapcrypto.dll).

７ Note

Before you attempt to refresh a SAP BW-based report that uses Kerberos SSO,
complete both the steps in this article and the steps in Configure Kerberos SSO.
Using CommonCryptoLib as your SNC library enables SSO connections to both SAP
BW Application Servers and SAP BW Message Servers.

７ Note

Configuring both libraries(sapcrypto and gx64krb5) on the same gateway server is

an unsupported scenario. It's not recommended to configure both libraries on the
same gateway server as it'll lead to a mix of libraries. If you want to use both
libraries, please fully separate the gateway server. For example, configure gx64krb5
for server A then sapcrypto for server B. Please remember that any failure on server
A which uses gx64krb5 is not supported, as gx64krb5 is no longer supported by
SAP and Microsoft.

Configure SAP BW to enable SSO using

CommonCryptoLib

７ Note

The on-premises data gateway is 64-bit software and therefore requires the 64-bit
version of CommonCryptoLib (sapcrypto.dll) to perform BW SSO. If you plan to test
the SSO connection to your SAP BW server in SAP GUI prior to attempting an SSO
connection through the gateway (recommended), you'll also need the 32-bit
version of CommonCryptoLib, as SAP GUI is 32-bit software.
1. Ensure that your BW server is correctly configured for Kerberos SSO using
CommonCryptoLib. If it is, you can use SSO to access your BW server (either
directly or through an SAP BW Message Server) with an SAP tool like SAP GUI that
has been configured to use CommonCryptoLib.

For more information on setup steps, see SAP Single Sign-On: Authenticate with
Kerberos/SPNEGO . Your BW server should use CommonCryptoLib as its SNC
Library and have an SNC name that starts with CN=, such as CN=BW1. For more
information on SNC name requirements (specifically, the snc/identity/as
parameter), see SNC Parameters for Kerberos Configuration .

2. If you haven't already done so, install the x64-version of the SAP .NET Connector
on the computer the gateway has been installed on.

You can check whether the component has been installed by attempting to
connect to your BW server in Power BI Desktop from the gateway computer. If you
can't connect by using the 2.0 implementation, the .NET Connector isn't installed
or hasn't been installed to the GAC.

3. Ensure that SAP Secure Login Client (SLC) isn't running on the computer the
gateway is installed on.

SLC caches Kerberos tickets in a way that can interfere with the gateway's ability to
use Kerberos for SSO.

4. If SLC is installed, uninstall it or make sure you exit SAP Secure Login Client. Right-
click the icon in the system tray and select Log Out and Exit before you attempt an
SSO connection by using the gateway.

SLC isn't supported for use on Windows Server machines. For more information,
see SAP Note 2780475 (s-user required).
5. If you uninstall SLC or select Log Out and Exit, open a cmd window and enter
klist purge to clear any cached Kerberos tickets before you attempt an SSO
connection through the gateway.

6. Download 64-bit CommonCryptoLib (sapcrypto.dll) version 8.5.25 or greater from

the SAP Launchpad, and copy it to a folder on your gateway machine. In the same
directory where you copied sapcrypto.dll, create a file named sapcrypto.ini, with
the following content:

ccl/snc/enable_kerberos_in_client_role = 1

The .ini file contains configuration information required by CommonCryptoLib to

enable SSO in the gateway scenario.

７ Note

These files must be stored in the same location; in other words,

/path/to/sapcrypto/ should contain both sapcrypto.ini and sapcrypto.dll.

Both the gateway service user and the Active Directory (AD) user that the service
user impersonates need read and execute permissions for both files. We
recommend granting permissions on both the .ini and .dll files to the
Authenticated Users group. For testing purposes, you can also explicitly grant
these permissions to both the gateway service user and the Active Directory user
you use for testing. In the following screenshot we've granted the Authenticated
Users group Read & execute permissions for sapcrypto.dll:
7. If you don't already have an SAP BW data source associated with the gateway you
want the SSO connection to flow through, add one on the Manage gateways page
in the Power BI service. If you already have such a data source, edit it:

Choose SAP Business Warehouse as the Data Source Type if you want to
create an SSO connection to a BW Application Server.
Select Sap Business Warehouse Message Server if you want to create an SSO
connection to a BW Message Server.

8. For SNC Library, select either the SNC_LIB or SNC_LIB_64 environment variable, or
Custom.

If you select SNC_LIB, you must set the value of the SNC_LIB_64 environment
variable on the gateway machine to the absolute path of the 64-bit copy of
 sapcrypto.dll on the gateway machine. For example,
C:\Users\Test\Desktop\sapcrypto.dll.

If you choose Custom, paste the absolute path to sapcrypto.dll into the
Custom SNC Library Path field that appears on the Manage gateways page.

9. For SNC Partner Name, enter the SNC Name of the BW server. Under Advanced
settings, ensure that Use SSO via Kerberos for DirectQuery queries is checked. Fill
in the other fields as if you were establishing a Windows Authentication
connection from PBI Desktop.

10. Create a CCL_PROFILE system environment variable and set its value to the path to
sapcrypto.ini.

The sapcrypto .dll and .ini files must exist in the same location. In the above
example, sapcrypto.ini and sapcrypto.dll are both located on the desktop.

11. Restart the gateway service.

 12. Run a Power BI report

Troubleshooting
If you're unable to refresh the report in the Power BI service, you can use gateway
tracing, CPIC tracing, and CommonCryptoLib tracing to diagnose the issue. Because
CPIC tracing and CommonCryptoLib are SAP products, Microsoft can't provide support
for them.

Gateway logs
1. Reproduce the issue.

2. Open the gateway app, and select Export logs from the Diagnostics tab.
CPIC tracing
1. To enable CPIC tracing, set two environment variables: CPIC_TRACE and
CPIC_TRACE_DIR.

The first variable sets the trace level and the second variable sets the trace file
directory. The directory must be a location that members of the Authenticated
Users group can write to.

2. Set CPIC_TRACE to 3 and CPIC_TRACE_DIR to whichever directory you want the

trace files written to. For example:

3. Reproduce the issue and ensure that CPIC_TRACE_DIR contains trace files.

CPIC tracing can diagnose higher level issues such as a failure to load the
sapcrypto.dll library. For example, here is a snippet from a CPIC trace file where a
.dll load error occurred:

Output

[Thr 7228] *** ERROR => DlLoadLib()==DLENOACCESS -

LoadLibrary("C:\Users\test\Desktop\sapcrypto.dll")
Error 5 = "Access is denied." [dlnt.c 255]
 If you encounter such a failure but you've set the Read & Execute permissions on
sapcrypto.dll and sapcrypto.ini as described in the section above, try setting the
same Read & Execute permissions on the folder that contains the files.

If you're still unable to load the .dll, try turning on auditing for the file. Examining
the resulting audit logs in the Windows Event Viewer might help you determine
why the file is failing to load. Look for a failure entry initiated by the impersonated
Active Directory user. For example, for the impersonated user MYDOMAIN\mytestuser
a failure in the audit log would look something like this:

Output

A handle to an object was requested.

Subject:
Security ID: MYDOMAIN\mytestuser
Account Name: mytestuser
Account Domain: MYDOMAIN
Logon ID: 0xCF23A8

Object:
Object Server: Security
Object Type: File
Object Name: <path information>\sapcrypto.dll
Handle ID: 0x0
Resource Attributes: -

Process Information:
Process ID: 0x2b4c
Process Name: C:\Program Files\On-premises data
gateway\Microsoft.Mashup.Container.NetFX45.exe

Access Request Information:

Transaction ID: {00000000-0000-0000-0000-000000000000}
Accesses: ReadAttributes

Access Reasons: ReadAttributes: Not granted

Access Mask: 0x80

Privileges Used for Access Check: -
Restricted SID Count: 0

CommonCryptoLib tracing
1. Turn on CommonCryptoLib tracing by adding these lines to the sapcrypto.ini file
you created earlier:
 ccl/trace/level=5
ccl/trace/directory=<drive>:\logs\sectrace

2. Change the ccl/trace/directory option to a location to which members of the

Authenticated Users group can write.

3. Alternatively, create a new .ini file to change this behavior. In the same directory as
sapcrypto.ini and sapcrypto.dll, create a file named sectrace.ini, with the following
content. Replace the DIRECTORY option with a location on your machine that
members of the Authenticated Users group can write to:

LEVEL = 5
DIRECTORY = <drive>:\logs\sectrace

4. Reproduce the issue and verify that the location pointed to by DIRECTORY
contains trace files.

5. When you're finished, turn off CPIC and CCL tracing.

For more information on CommonCryptoLib tracing, see SAP Note 2491573 (SAP
s-user required).

Impersonation
This section describes troubleshooting symptoms and resolution steps for
impersonation issues.

Symptom: When looking at the GatewayInfo[date].log you find an entry similar to the
following: About to impersonate user DOMAIN\User (IsAuthenticated: True,
ImpersonationLevel: Impersonation). If the value for ImpersonationLevel is different
from Impersonation, impersonation is not happening properly.

Resolution: Follow the steps found in grant the gateway service account local policy
rights on the gateway machine article. Restart the gateway service after changing the
configuration.

Validation: Refresh or create the report and collect the GatewayInfo[date].log. Open the
latest GatewayInfo log file and check again the following string: About to impersonate
user DOMAIN\User (IsAuthenticated: True, ImpersonationLevel: Impersonation) to
ensure that the value for ImpersonationLevel matches Impersonation.
Delegation
Delegation issues usually appear in the Power BI service as generic errors. To determine
whether delegation is the issue, it's useful to collect the Wireshark traces and use
Kerberos as a filter. For Kerberos errors reference, consult the blog post. The rest of this
section describes troubleshooting symptoms and resolution steps for delegation issues.

Symptom: In the Power BI service you may encounter an unexpected error, similar to the
following screenshot. the GatewayInfo[date].log you'll see [DM.GatewayCore] ingesting
an exception during Ado query execution attempt for clientPipelineId and the import
[0D_NW_CHANN] matches no exports.

In the Mashup[date].log you see the generic error GSS-API(maj): No credentials were
supplied.

Looking into the CPIC traces (sec-Microsoft.Mashup*.trc) you will see something similar
to the following:

[Thr 4896] *** ERROR => SncPEstablishContext() failed for target='p:CN=BW5'

[sncxxall.c 3638]
[Thr 4896] *** ERROR => SncPEstablishContext()==SNCERR_GSSAPI [sncxxall.c
3604]
[Thr 4896] GSS-API(maj): No credentials were supplied
[Thr 4896] Unable to establish the security context
[Thr 4896] target="p:CN=BW5"
[Thr 4896] <<- SncProcessOutput()==SNCERR_GSSAPI
[Thr 4896]
[Thr 4896] LOCATION CPIC (TCP/IP) on local host HNCL2 with Unicode
[Thr 4896] ERROR GSS-API(maj): No credentials were supplied
[Thr 4896] Unable to establish the security context
[Thr 4896] target="p:CN=BW5"
[Thr 4896] TIME Thu Oct 15 20:49:31 2020
[Thr 4896] RELEASE 721
[Thr 4896] COMPONENT SNC (Secure Network Communication)
[Thr 4896] VERSION 6
[Thr 4896] RC -4
[Thr 4896] MODULE sncxxall.c
 [Thr 4896] LINE 3604
[Thr 4896] DETAIL SncPEstablishContext
[Thr 4896] SYSTEM CALL gss_init_sec_context
[Thr 4896] COUNTER 3
[Thr 4896]
[Thr 4896] *** ERROR => STISEND:STISncOut failed 20 [r3cpic.c 9834]
[Thr 4896] STISearchConv: found conv without search

The error becomes clearer in the sectraces from the Gateway machine sec-
Microsoft.Mashup.Con-[].trc:

[2020.10.15 20:31:38.396000][4][Microsoft.Mashup.Con][Kerberos ][ 3616]

AcquireCredentialsHandleA called successfully.
[2020.10.15 20:31:38.396000][2][Microsoft.Mashup.Con][Kerberos ][ 3616]
InitializeSecurityContextA returned -2146893053 (0x80090303). Preparation
for kerberos failed!
[2020.10.15 20:31:38.396000][2][Microsoft.Mashup.Con][Kerberos ][ 3616]
Getting kerberos ticket for 'SAP/BW5' failed (user name is
[email protected])
[2020.10.15 20:31:38.396000][2][Microsoft.Mashup.Con][Kerberos ][ 3616]
Error for requested algorithm 18: 0/C000018B The security database on the
server does not have a computer account for this workstation trust
relationship.
[2020.10.15 20:31:38.396000][2][Microsoft.Mashup.Con][Kerberos ][ 3616]
Error for requested algorithm 17: 0/C000018B The security database on the
server does not have a computer account for this workstation trust
relationship.
[2020.10.15 20:31:38.396000][2][Microsoft.Mashup.Con][Kerberos ][ 3616]
Error for requested algorithm 23: 0/C000018B The security database on the
server does not have a computer account for this workstation trust
relationship.
[2020.10.15 20:31:38.396000][2][Microsoft.Mashup.Con][Kerberos ][ 3616]
Error for requested algorithm 3: 0/C000018B The security database on the
server does not have a computer account for this workstation trust
relationship.

You can also see the issue if you look at WireShark traces.
 ７ Note

The other errors KRB5KDC_ERR_PREAUTH_REQUIRED can be safely ignored.

Resolution: You must add an SPN SAP/BW5 to a service account. Detailed information
and steps are available in the SAP documentation .

You may run into a similar, but not identical error that manifests in WireShark traces as
the following error KRB5KDC_ERR_BADOPTION:

This error indicates the SPN SAP/BW5 could be found, but it's not in the Services to
which this account can present delegated credentials at the Delegation tab from the
Gateway service account. To fix this issue, follow the steps to configure the gateway
service account for standard kerberos constrained delegation.
Validation: Proper configuration will prevent generic or unexpected errors to be
presented by the gateway. If you still see errors, check the configuration of the gateway
itself, or the configuration of the BW server.

Credentials errors
This section describes troubleshooting symptoms and resolution steps for credentials
error issues. You may also see generic errors from the Power BI service, as described in
the earlier section on delegation.

There are different resolutions, based on the symptoms you see in the data source (SAP
BW), so we'll review both.

Symptom 1: In the sectraces sec-disp+work[].trc from the BW Server, you see traces
similar to the following:

[2020.05.26 14:21:28.668325][4][disp+work ][SAPCRYPTOLIB][435584] {

gss_display_name [2020.05.26 14:21:28.668338][4][disp+work ][GSS ][435584]
gss_display_name output buffer (41 bytes) [2020.05.26 14:21:28.668338][4]
[disp+work ][GSS ][435584] [email protected]@CONTOSO.COM

Resolution: Complete the configuration steps to set user mapping configuration

parameters on the gateway machine if necessary. You'll need to complete those steps
even if you already have the Azure AD Connect configured.

Validation: You'll be able to successfully load the report in the Power BI service. If not
successful, see the steps in symptom 2.

Symptom 2: In the sectraces sec-disp+work[].trc from the BW Server, you see traces
similar to the following:

[2020.10.19 23:10:15.469000][4][disp+work.EXE ][SAPCRYPTOLIB][ 4460] {

gss_display_name
[2020.10.19 23:10:15.469000][4][disp+work.EXE ][GSS ][ 4460]
gss_display_name output buffer (23 bytes)
[2020.10.19 23:10:15.469000][4][disp+work.EXE ][GSS ][ 4460]
[email protected]

Resolution: Check whether the Kerberos external ID for the User match what the
sectraces are showing.
 1. Open SAP Logon.
2. Use the SU01 transaction.
3. Edit the user.
4. Navigate to the SNC tab, verify that the SNC name matches what is shown in your
logs.

Validation: When properly completed, you'll be able to create and refresh reports in the
Power BI service.

Next steps
For more information about the on-premises data gateway and DirectQuery, see the
following resources:

What is an on-premises data gateway?

DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP BW
DirectQuery and SAP HANA
Use Kerberos for single sign-on (SSO) to
SAP BW using gx64krb5
Article • 03/29/2022

This article describes how to configure your SAP BW data source to enable SSO from the
Power BI service by using gx64krb5.

） Important

Microsoft will allow you to create connections using SNC libraries (like gx64krb5)
but will not provide support for these configurations. Additionally SAP no longer
supports the gx64krb5 for on-premises data gateways in Power BI and the steps
required to configure it for the gateway are significantly more complex compared
to CommonCryptoLib. As a result, Microsoft recommends using CommonCryptoLib
instead. For more information, see SAP Note 352295 . Note that gx64krb5
doesn't allow for SSO connections from the data gateway to SAP BW Message
Servers; only connections to SAP BW Application Servers are possible. This
restriction doesn't exist if you use CommonCryptoLib as your SNC library. For
information about how to configure SSO by using CommonCryptoLib, see
Configure SAP BW for SSO using CommonCryptoLib. Use CommonCryptoLib or
gx64krb5 as your SNC library, but not both. Do not complete the configuration
steps for both libraries.

７ Note

Configuring both libraries(sapcrypto and gx64krb5) on the same gateway server is

an unsupported scenario. It’s not recommended to configure both libraries on the
same gateway server as it’ll lead to a mix of libraries. If you want to use both
libraries, please fully separate the gateway server. For example, configure gx64krb5
for server A then sapcrypto for server B. Please remember that any failure on server
A which uses gx64krb5 is not supported, as gx64krb5 is no longer supported by
SAP and Microsoft.

This guide is comprehensive; if you've already completed some of the described steps,
you can skip them. For example, you might have already configured your SAP BW server
for SSO using gx64krb5.
Set up gx64krb5 on the gateway machine and
the SAP BW server
The gx64krb5 library must be used by both the client and server to complete an SSO
connection through the gateway. That is, both the client and server must be using the
same SNC library.

1. Download gx64krb5.dll from SAP Note 2115486 (SAP s-user required). Ensure
you have at least version 1.0.11.x. Also, download gsskrb5.dll (the 32-bit version of
the library) if you want to test the SSO connection in SAP GUI before you attempt
the SSO connection through the gateway (recommended). The 32-bit version is
required to test with SAP GUI because SAP GUI is 32-bit only.

2. Put gx64krb5.dll in a location on your gateway machine that's accessible by your

gateway service user. If you want to test the SSO connection with SAP GUI, also put
a copy of gsskrb5.dll on your machine and set the SNC_LIB environment variable
to point to it. Both the gateway service user and the Active Directory (AD) users
that the service user will impersonate need read and execute permissions for the
copy of gx64krb5.dll. We recommend granting permissions on the .dll to the
Authenticated Users group. For testing purposes, you can also explicitly grant
these permissions to both the gateway service user and the Active Directory user
you use to test.

3. If your BW server hasn't already been configured for SSO using gx64krb5.dll, put
another copy of the .dll on your SAP BW server machine in a location accessible by
the SAP BW server.

For more information on configuring gx64krb5.dll for use with an SAP BW server,
see SAP documentation (SAP s-user required).

4. On the client and server machines, set the SNC_LIB and SNC_LIB_64 environment
variables:

If you use gsskrb5.dll, set the SNC_LIB variable to its absolute path.
If you use gx64krb5.dll, set the SNC_LIB_64 variable to its absolute path.

Configure an SAP BW service user and enable

SNC communication on the BW server
Complete this section if you haven't already configured your SAP BW server for SNC
communication (for example, SSO) by using gx64krb5.
７ Note

This section assumes that you've already created a service user for BW and bound a
suitable SPN to it (that is, a name that begins with SAP/).

1. Give the service user access to your SAP BW Application Server:

a. On the SAP BW server machine, add the service user to the Local Admin group.
Open the Computer Management program and identify the Local Admin group
for your server.

b. Double-click the Local Admin group, and select Add to add your service user to
the group.

c. Select Check Names to ensure you've entered the name correctly, and then
select OK.

2. Set the SAP BW server's service user as the user that starts the SAP BW server
service on the SAP BW server machine:

a. Open Run, and then enter Services.msc.

 b. Find the service corresponding to your SAP BW Application Server instance,
right-click it, and then select Properties.

c. Switch to the Log on tab, and change the user to your SAP BW service user.

d. Enter the user's password, and then select OK.

3. In SAP Logon, sign in to your server and set the following profile parameters by
using the RZ10 transaction:

a. Set the snc/identity/as profile parameter to p:<SAP BW service user you

created>. For example, p:[email protected]. Note that p:
precedes the service user's UPN, as opposed to p:CN=, which precedes the UPN
when you use CommonCryptoLib as the SNC library.

b. Set the snc/gssapi_lib profile parameter to <path to gx64krb5.dll on the BW

server>. Place the library in a location that the SAP BW Application Server can
access.

c. Set the following additional profile parameters, changing the values as required
to fit your needs. The last five options enable clients to connect to the SAP BW
server by using SAP Logon without having SNC configured.

Setting Value

snc/data_protection/max 3

snc/data_protection/min 1

snc/data_protection/use 9

snc/accept_insecure_cpic 1

snc/accept_insecure_gui 1
 Setting Value

snc/accept_insecure_r3int_rfc 1

snc/accept_insecure_rfc 1

snc/permit_insecure_start 1

d. Set the snc/enable property to 1.

4. After you set these profile parameters, open the SAP Management Console on the
server machine and restart the SAP BW instance.

If the server won't start, confirm that you've set the profile parameters correctly.
For more information on profile parameter settings, see the SAP documentation .
You can also consult the Troubleshooting section in this article.

Map an SAP BW user to an Active Directory

user
If you haven't done so already, map an Active Directory user to an SAP BW Application
Server user and test the SSO connection in SAP Logon.

1. Sign in to your SAP BW server with SAP Logon. Run transaction SU01.

2. For User, enter the SAP BW user for which you want to enable SSO connection.
Select the Edit icon (pen icon) near the top left of the SAP Logon window.
3. Select the SNC tab. In the SNC name input box, enter p:<your Active Directory
user>@<your domain>. For the SNC name, p: must precede the Active Directory
user's UPN. Note that the UPN is case-sensitive.

The Active Directory user you specify should belong to the person or organization
for whom you want to enable SSO access to the SAP BW Application Server. For
example, if you want to enable SSO access for the user
[email protected], enter p:[email protected].
 4. Select the Save icon (floppy disk image) near the top left of the screen.

Test sign in via SSO

Verify that you can sign in to the server by using SAP Logon through SSO as the Active
Directory user for whom you've enabled SSO access:

1. As the Active Directory user for which you've just enabled SSO access, sign in to a
machine in your domain on which SAP Logon is installed. Launch SAP Logon, and
create a new connection.

2. Copy the gsskrb5.dll file you downloaded earlier to a location on the machine you
signed-in to. Set the SNC_LIB environment variable to the absolute path of this
location.

3. Launch SAP Logon, and create a new connection.

4. In the Create New System Entry screen, select User Specified System, then select
Next.
5. Fill in the appropriate details on the next screen, including the application server,
instance number, and system ID. Then, select Finish.

6. Right-click the new connection, select Properties, and then select the Network tab.

7. In the SNC Name box, enter p:<SAP BW service user's UPN>. For example,
p:[email protected]. Select OK.
 8. Double-click the connection you just created to attempt an SSO connection to
your SAP BW server.

If this connection succeeds, continue to the next section. Otherwise, review the
earlier steps in this document to make sure they've been completed correctly, or
review the Troubleshooting section. If you can't connect to the SAP BW server via
SSO in this context, you won't be able to connect to the SAP BW server by using
SSO in the gateway context.

Add registry entries to the gateway machine

Add required registry entries to the registry of the machine that the gateway is installed
on, and to machines intended to connect from Power BI Desktop. To add these registry
entries, run the following commands:

REG ADD HKLM\SOFTWARE\Wow6432Node\SAP\gsskrb5 /v ForceIniCredOK /t REG_DWORD

/d 1 /f

REG ADD HKLM\SOFTWARE\SAP\gsskrb5 /v ForceIniCredOK /t REG_DWORD /d 1 /f

Add a new SAP BW Application Server data
source to the Power BI service, or edit an
existing one
1. In the data source configuration window, enter the SAP BW Application Server's
Hostname, System Number, and client ID, as you would to sign in to your SAP BW
server from Power BI Desktop.

2. In the SNC Partner Name field, enter p:<SPN you mapped to your SAP BW service
user>. For example, if the SPN is SAP/[email protected], enter
p:SAP/[email protected] in the SNC Partner Name field.

3. For the SNC Library, select the Custom option and provide the absolute path for
GX64KRB5.DLL or GSSKRB5.DLL on the gateway machine.

4. Select Use SSO via Kerberos for DirectQuery queries, and then select Apply. If the
test connection is not successful, verify that the previous setup and configuration
steps were completed correctly.

5. Run a Power BI report

Troubleshooting

Troubleshoot gx64krb5 configuration

If you encounter any of the following problems, follow these steps to troubleshoot the
gx64krb5 installation and SSO connections:

You encounter errors when you complete the gx64krb5 setup steps. For example,
the SAP BW server won't start after you've changed the profile parameters. View
the server logs (…work\dev_w0 on the server machine) to troubleshoot these
errors.

You can't start the SAP BW service due to a sign-on failure. You might have
provided the wrong password when you set the SAP BW start-as user. Verify the
password by signing in as the SAP BW service user on a machine in your Active
Directory environment.

You get errors about underlying data source credentials (for example, SQL Server),
which prevent the server from starting, verify that you've granted the service user
access to the SAP BW database.
 You get the following message: (GSS-API) specified target is unknown or
unreachable. This error usually means you have the wrong SNC name specified.
Make sure to use p: only, not p:CN=, to precede the service user's UPN in the client
application.

You get the following message: (GSS-API) An invalid name was supplied. Make sure
p: is the value of the server's SNC identity profile parameter.

You get the following message: (SNC error) the specified module could not be found.
This error is often caused by placing gx64krb5.dll in a location that requires
elevated privileges (../administrator rights) to access.

Troubleshoot gateway connectivity issues

1. Check the gateway logs. Open the gateway configuration application, and select
Diagnostics, then Export logs. The most recent errors are at the end of any log
files you examine.

2. Turn on SAP BW tracing, and review the generated log files. There are several
different types of SAP BW tracing available (for example, CPIC tracing):
 a. To enable CPIC tracing, set two environment variables: CPIC_TRACE and
CPIC_TRACE_DIR.

The first variable sets the trace level and the second variable sets the trace file
directory. The directory must be a location that members of the Authenticated
Users group can write to.

b. Set CPIC_TRACE to 3 and CPIC_TRACE_DIR to whichever directory you want the

trace files written to. For example:

c. Reproduce the issue and ensure that CPIC_TRACE_DIR contains trace files.

d. Examine the contents of the trace files to determine the blocking issue. For
example, you may find that gx64krb5.dll was not loaded properly, or that an Active
Directory user different than the one you were expecting initiated the SSO
connection attempt.

Next steps
For more information about the on-premises data gateway and DirectQuery, see the
following resources:

What is an on-premises data gateway?

DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP BW
DirectQuery and SAP HANA
Use Kerberos for SSO to Teradata
Article • 02/23/2023

This article describes a specific added requirement to successfully enable single sign-on
(SSO) to Teradata from the Power BI service.

If Teradata identifies user accounts by using sAMAccountNames, you must set

FullDomainResolutionEnabled on the gateway to True .

If Teradata identifies user accounts by using User Principal Names (UPNs), keep
FullDomainResolutionEnabled on the gateway set to False .

Enable SSO for Teradata

To change the FullDomainResolutionEnabled configuration on the gateway to enable
SSO for Teradata:

1. In the on-premises gateway directory at %ProgramFiles%\On-premises data

gateway, open the configuration file
Microsoft.PowerBI.DataMovement.Pipeline.GatewayCore.dll.config.

2. In the file, find the FullDomainResolutionEnabled property and change its value to
True .

XML

<setting name=" FullDomainResolutionEnabled " serializeAs="String">

<value>True</value>
</setting>

Next steps
For more information about the on-premises data gateway and DirectQuery, see the
following resources:

What is an on-premises data gateway?

DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP Business Warehouse (BW)
DirectQuery and SAP HANA
Use Security Assertion Markup
Language for SSO from Power BI to on-
premises data sources
Article • 05/17/2022

By enabling single sign-on (SSO), you can make it easy for Power BI reports and
dashboards to refresh data from on-premises sources while you respect user-level
permissions that are configured on those sources. To enable seamless SSO connectivity,
you use Security Assertion Markup Language (SAML) .

７ Note

You can connect to only one data source using Single Sign-On SAML with an on-
premises data gateway. To connect to an additional data source using Single Sign-
On SAML, you must use a different on-premises data gateway.

Supported data sources for SAML

Microsoft currently supports SAP HANA with SAML. For more information about setting
up and configuring single sign-on for SAP HANA by using SAML, see SAML SSO for BI
Platform to HANA .

We support additional data sources with Kerberos (including SAP HANA).

For SAP HANA, we recommend that you enable encryption before you establish a SAML
SSO connection. To enable encryption, configure the HANA server to accept encrypted
connections, and then configure the gateway to use encryption to communicate with
your HANA server. Because the HANA ODBC driver doesn't encrypt SAML assertions by
default, the signed SAML assertion is sent from the gateway to the HANA server in the
clear and is vulnerable to interception and reuse by third parties.

） Important

Because SAP no longer supports OpenSSL , Microsoft has also discontinued its
support. Your existing connections continue to work but you can no longer create
new connections. Use SAP Cryptographic Library (CommonCryptoLib), or sapcrypto,
instead.
Configure the gateway and data source
To use SAML, you must establish a trust relationship between the HANA servers for
which you want to enable SSO and the gateway. In this scenario, the gateway serves as
the SAML identity provider (IdP). You can establish this relationship in various ways. SAP
recommends that you use CommonCryptoLib to complete the setup steps. For more
information, see the official SAP documentation.

Create the certificates

You can establish a trust relationship between a HANA server and the gateway IdP by
signing the gateway IdP's X509 certificate with a root certificate authority (CA) that's
trusted by the HANA server.

To create the certificates, do the following:

1. On the device that's running SAP HANA, create an empty folder to store your
certificates, and then go to that folder.

2. Create the root certificates by running the following command:

openssl req -new -x509 -newkey rsa:2048 -days 3650 -sha256 -keyout
CA_Key.pem -out CA_Cert.pem -extensions v3_ca'''

Be sure to copy and save the passphrase to use this certificate to sign other
certificates. You should see the CA_Cert.pem and CA_Key.pem files being created.

3. Create the IdP certificates by running the following command:

openssl req -newkey rsa:2048 -days 365 -sha256 -keyout IdP_Key.pem -out
IdP_Req.pem -nodes

You should see the IdP_Key.pem and IdP_Req.pem files being created.

4. Sign the IdP certificates with the root certificates:

openssl x509 -req -days 365 -in IdP_Req.pem -sha256 -extensions

usr_cert -CA CA_Cert.pem -CAkey CA_Key.pem -CAcreateserial -out
IdP_Cert.pem
 You should see the CA_Cert.srl and IdP_Cert.pem files being created. At this time,
you're concerned only with the IdP_Cert.pem file.

Create mapping for the SAML identity provider certificate

To create mapping for the SAML Identity Provider certificate, do the following:

1. In SAP HANA Studio, right-click your SAP HANA server name, and then select
Security > Open Security Console > SAML Identity Provider.

2. Select the SAP Cryptographic Library option. Do not use the OpenSSL
Cryptographic Library option, which is deprecated by SAP.

3. To import the signed certificate IdP_Cert.pem, select the blue Import button, as
shown in the following image:

4. Remember to assign a name for your identity provider.

Import and create the signed certificates in HANA

To import and create the signed certificates in HANA, do the following:

1. In SAP HANA Studio, run the following query:

 CREATE CERTIFICATE FROM '<idp_cert_pem_certificate_content>'

Here's an example:

CREATE CERTIFICATE FROM

'-----BEGIN CERTIFICATE-----
MIIDyDCCArCgA...veryLongString...0WkC5deeawTyMje6
-----END CERTIFICATE-----
'

2. If there's no personal security environment (PSE) with purpose SAML, create one
by running the following query in SAP HANA Studio:

CREATE PSE SAMLCOLLECTION;

set pse SAMLCOLLECTION purpose SAML;

3. Add the newly created signed certificate to the PSE by running the following
command:

alter pse SAMLCOLLECTION add CERTIFICATE <certificate_id>;

For example:

alter pse SAMLCOLLECTION add CERTIFICATE 1978320;

You can check the list of created certificates by running the following query:

select * from PUBLIC"."CERTIFICATES"

The certificate is now properly installed. To confirm the installation, you can run the
following query:
 select * from "PUBLIC"."PSE_CERTIFICATES"

Map the user

To map the user, do the following:

1. In SAP HANA Studio, select the Security folder.

2. Expand Users, and then select the user that you want to map your Power BI user
to.

3. Select the SAML checkbox, and then select Configure, as shown in the following
image:

4. Select the identity provider that you created in the Create mapping for the SAML
identity provider certificate section. For External Identity, enter the Power BI user's
UPN (ordinarily, the email address the user uses to sign in to Power BI), and then
select Add.

If you've configured your gateway to use the ADUserNameReplacementProperty

configuration option, enter the value that will replace the Power BI user's original
UPN. For example, if you set ADUserNameReplacementProperty to
SAMAccountName, enter the user's SAMAccountName.
Configure the gateway
Now that you've configured the gateway certificate and identity, convert the certificate
to a PFX file format, and then configure the gateway to use the certificate by doing the
following:

1. Convert the certificate to PFX format by running the following command. This
command names the resulting file samlcert.pfx and sets root as its password, as
shown here:

openssl pkcs12 -export -out samltest.pfx -in IdP_Cert.pem -inkey

IdP_Key.pem -passin pass:root -passout pass:root

2. Copy the PFX file to the gateway machine:

a. Double-click samltest.pfx, and then select Local Machine > Next.

b. Enter the password, and then select Next.

c. Select Place all certificates in the following store, and then select Browse >
Personal > OK.

d. Select Next, and then select Finish.

3. To grant the gateway service account access to the private key of the certificate, do
the following:

a. On the gateway machine, run Microsoft Management Console (MMC).

b. In MMC, select File > Add/Remove Snap-in.

c. Select Certificates > Add, and then select Computer account > Next.

d. Select Local Computer > Finish > OK.

e. Expand Certificates > Personal > Certificates, and then look for the certificate.

f. Right-click the certificate, and then select All Tasks > Manage Private Keys.

g. Add the gateway service account to the list. By default, the account is NT
SERVICE\PBIEgwService. You can find out which account is running the gateway
 service by running services.msc and then looking for On-premises data gateway
service.

Finally, add the certificate thumbprint to the gateway configuration:

1. To list the certificates on your machine, run the following PowerShell command:

PowerShell

Get-ChildItem -path cert:\LocalMachine\My

2. Copy the thumbprint for the certificate you created.

3. Go to the gateway directory, which is C:\Program Files\On-premises data gateway

by default.

4. Open PowerBI.DataMovement.Pipeline.GatewayCore.dll.config, and then look for

the SapHanaSAMLCertThumbprint section. Paste the thumbprint you copied in
step 2.

5. Restart the gateway service.

Run a Power BI report

Now you can use the Manage Gateway page in Power BI to configure the SAP HANA
data source. Under Advanced Settings, enable SSO via SAML. By doing so, you can
publish reports and datasets binding to that data source.
 ７ Note

SSO uses Windows Authentication so make sure the windows account can access
the gateway machine. If not sure, make sure to add NT-AUTHORITY\Authenticated
Users (S-1-5-11) to the local machine “Users” group.

Troubleshoot using SAML for single sign-on to

SAP HANA
This section provides extensive steps to troubleshoot using SAML for single sign-on to
SAP HANA. Using these steps can help you self-diagnose and correct any issues you
might face.

Rejected credentials
After you configure SAML-based SSO, you might see the following error in the Power BI
portal: "The credentials provided cannot be used for the SapHana source." This error
indicates that the SAML credentials were rejected by SAP HANA.

Server-side authentication traces provide detailed information for troubleshooting

credential issues on SAP HANA. To configure tracing for your SAP HANA server, do the
following:

1. On the SAP HANA server, turn on the authentication trace by running the following
query:

ALTER SYSTEM ALTER CONFIGURATION ('indexserver.ini', 'SYSTEM') set

('trace', 'authentication') = 'debug' with reconfigure

2. Reproduce the issue.

3. In SAP HANA Studio, open the administration console, and then select the
Diagnosis Files tab.

4. Open the most recent index server trace, and then search for
SAMLAuthenticator.cpp.

You should find a detailed error message that indicates the root cause, as shown in
the following example:
 [3957]{-1}[-1/-1] 2018-09-11 21:40:23.815797 d Authentication
SAMLAuthenticator.cpp(00091) : Element
'{urn:oasis:names:tc:SAML:2.0:assertion}Assertion', attribute 'ID':
'123123123123123' is not a valid value of the atomic type 'xs:ID'.
[3957]{-1}[-1/-1] 2018-09-11 21:40:23.815914 i Authentication
SAMLAuthenticator.cpp(00403) : No valid SAML Assertion or SAML Protocol
detected

5. After you've finished troubleshooting, turn off the authentication trace by running
the following query:

ALTER SYSTEM ALTER CONFIGURATION ('indexserver.ini', 'SYSTEM') UNSET

('trace', 'authentication');

Verify and troubleshoot gateway errors

To follow the procedures in this section, you need to collect gateway logs.

SSL error (certificate)

Error symptoms

This issue has multiple symptoms. When you try to add a new data source, you might
see an error message like the following:

Unable to connect: We encountered an error while trying to connect to . Details:

"We could not register this data source for any gateway instances within this

cluster. Please find more details below about specific errors for each gateway
instance."

When you try to create or refresh a report, you might see an error message like the one
in the following image:
When you investigate the Mashup[date]*.log, you'll see the following error message:

A connection was successfully established with the server, but then an error

occurred during the login process and the certificate chain was issued by an
authority that is not trusted

Resolution

To resolve this SSL error, go to the data source connection and then, in the Validate
Server Certificate dropdown list, select No, as shown in the following image:

After you've selected this setting, the error message will no longer appear.

Gateway SignXML error

The gateway SignXML error can be the result of incorrect SapHanaSAMLCertThumbprint
settings, or it can be an issue with the HANA server. Entries in the gateway logs help
identify where the issue resides, and how to resolve it.

Error symptoms

Log entries for SignXML: Found the cert... : If your GatewayInfo[date].log file contains
this error, the SignXML cert was found, and your troubleshooting efforts should focus
on steps that are found in the "Verify and troubleshoot the HANA server side" section.

Log entries for Couldn't find saml cert : If your GatewayInfo[date].log file contains this
error, SapHanaSAMLCertThumbprint is set incorrectly. The following resolution section
describes how to resolve the issue.

Resolution

To properly set SapHanaSAMLCertThumbprint, follow the instructions in the "Configure

the gateway" section. The instructions begin with Finally, add the certificate thumbprint
to the gateway configuration.

After you've changed the configuration file, you need to restart the gateway service for
the change to take effect.

Validation

When SapHanaSAMLCertThumbprint is properly set, your gateway logs will have entries
that include SignXML: Found the cert... . At this point, you should be able to proceed
to the "Verify and troubleshoot the HANA server side" section.

If the gateway is unable to use the certificate to sign the SAML assertion, you might see
an error in the logs that's similar to the following:

GatewayPipelineErrorCode=DM_GWPipeline_UnknownError GatewayVersion=

InnerType=CryptographicException InnerMessage=<pi>Signing key is not loaded.</pi>

InnerToString=<pi>System.Security.Cryptography.CryptographicException: Signing key

is not loaded.

To resolve this error, follow the instructions beginning with step 3 in the "Configure the
gateway" section.

After you've changed the configuration, restart the gateway service for the change to
take effect.

Verify and troubleshoot the HANA server side

Use the solutions in this section if the gateway can find the certificate and sign the
SAML assertion but you're still experiencing errors. You'll need to collect HANA
authentication traces, as described earlier in the "Rejected credentials" section.

The SAML identity provider

The presence of the Found SAML provider string in the HANA authentication traces
indicates that the SAML identity provider is configured properly. If the string is not
present, the configuration is incorrect.

Resolution

First, determine whether your organization is using OpenSSL or commoncrypto as the

sslcryptoprovider. To determine which provider is being used, do the following:

1. Open SAP HANA Studio.

2. Open the Administration Console for the tenant that you're using.

3. Select the Configuration tab, and use sslcryptoprovider as a filter, as shown in the
following image:

Next, verify that the cryptographic library is set correctly by doing the following:

1. Go to Security Console in SAP HANA Studio by selecting the SAML Identity

Providers tab, and do either of the following:

If the sslcryptoprovider is OpenSSL, select OpenSSL Cryptographic Library.

If the sslcryptoprovider is commonCrypto, select SAP Cryptographic Library.

In the following image, SAP Cryptographic Library is selected:

 2. Deploy your changes by selecting the Deploy button at the upper right, as shown
in the following image:

Validation

When the traces are properly configured, they'll report Found SAML provider and will not
report SAML Provider not found . You can proceed to the next section, "Troubleshoot the
SAML assertion signature."

If the cryptographic provider is set but SAML Provider not found is still being reported,
search for a string in the trace that begins with the following text:

Search SAML provider for certificate with subject =

In that string, ensure that the subject and issuer are exactly the same as displayed in the
SAML identity provider tab in Security Console. A difference of even a single character
can cause the problem. If you find a difference, you can fix the issue in the SAP
Cryptographic Library so that the entries match exactly.

If changing the SAP Cryptographic Library doesn't fix the issue, you can manually edit
the Issued To and Issued By fields simply by double-clicking them.

Troubleshoot the SAML assertion signature

You might find HANA authentication traces that contain entries similar to the following:

[48163]{-1}[-1/-1] 2020-09-11 21:15:18.896165 i Authentication

SAMLAuthenticator.cpp(00398) : Unable to verify XML signature [48163]{-1}[-1/-1]

2020-09-11 21:15:18.896168 i Authentication MethodSAML.cpp(00103) : unsuccessful

login attempt with SAML ticket!

The presence of such entries means that the signature isn't trusted.

Resolution

If you're using OpenSSL as your sslcryptoprovider, check to see whether the trust.pem
and key.pem files are in the SSL directory. For more information, see the SAP blog
Securing the communication between SAP HANA Studio and SAP HANA Server through
SSL .

If you're using commoncrypto as your sslcryptoprovider, check to see whether there's a

collection with your certificate in the tenant.

Validation

When the traces are properly configured, they'll report Found valid XML signature .

Troubleshoot the UPN mapping

You might find HANA traces that contain entries similar to the following:

SAMLAuthenticator.cpp(00886) : Assertion Subject NameID: `[email protected]`

SAMLAuthenticator.cpp(00398) : Database user does not exist

The error indicates that nameId [email protected] is found in the SAML assertions, but
it doesn't exist or isn't mapped correctly in HANA Server.

Resolution

Go to the HANA database user and, under the selected SAML checkbox, select the
Configure link. The following window appears:

As the error message describes, HANA was trying to find [email protected], but the
external identity is displayed only as johnny. These two values must match. To resolve
the issue, under External Identity, change the value to [email protected]. Note that
this value is case sensitive.

Next steps
For more information about the on-premises data gateway and DirectQuery, see the
following resources:

What is an on-premises data gateway?

DirectQuery in Power BI
Data sources supported by DirectQuery
DirectQuery and SAP Business Warehouse (BW)
DirectQuery and SAP HANA
Test single sign-on (SSO) configuration
Article • 12/21/2023

Single sign-on (SSO) enables each Power BI user to access the precise data they have
permissions for in an underlying data source. Many Power BI data sources are enabled
for SSO, using either Kerberos constrained delegation or Security Assertion Markup
Language (SAML). For more information, see Overview of single sign-on for on-premises
data gateways in Power BI.

Setting up SSO is complex, so you can use the test single sign-on (SSO) configuration
feature to test your configuration.

The single sign-on test:

Lets the gateway connect to the data source by using a test User Principal Name
(UPN) that you provide.
Validates the SSO setup, which includes checking UPN mapping to a local Active
Directory (AD) identity for impersonation and data source access.
Helps identify problems if connection failures occur. For example, an error
message indicates if a UPN maps to a local AD identity that doesn't have access to
the data source.

The test single sign-on feature works for both Kerberos and SAML-based SSO for the
data sources listed in Supported data sources for SSO. For Kerberos constrained
delegation, the test single sign-on feature can help test SSO for both DirectQuery and
Import, or only DirectQuery data sources.

） Important

The test single sign-on feature requires the March 2021 gateway release or later.

Test SSO for the gateway

To test the SSO configuration:

1. From Manage connections and gateways in Power BI, select Settings for the data
source.
2. In the Settings pane, under Single sign-on, select Test single sign-on.

3. Provide a User Principal Name to test.

If the gateway cluster is able to impersonate the user and successfully connect to
the data source, the test succeeds, as shown in the following image:
Troubleshooting
This section describes common errors you might see when testing single sign-on, and
actions you can take to fix them.

Impersonation error
If the gateway cluster can't impersonate the user and connect to the data source, the
test fails with the error message: Error: The on-premises data gateway's service
account failed to impersonate the user.
There can be the following possible causes and solutions:

The user doesn't exist in Microsoft Entra ID. Check if the user is present in
Microsoft Entra ID.
The user isn't mapped correctly to a local AD account. Check configurations and
follow the steps in Overview of single sign-on for on-premises data gateways in
Power BI.
The gateway doesn't have impersonation rights. Grant the gateway service account
local policy rights on the gateway machine as described in Grant the gateway
service account local policy rights on the gateway machine.

Invalid credentials error

The error Error: Invalid connection credentials appears when the gateway can't connect
to the data source, because the provided UPN doesn't have access to the data source.
Check whether the data source has been misconfigured to deny access to the user. You
may need to work with your data source/database administrator to access the data
source's configuration and settings.

Next steps
Overview of single sign-on (SSO) for gateways in Power BI
Single sign-on (SSO) - Kerberos
Single sign-on (SSO) - SAML
Microsoft Entra SSO
Article • 01/23/2024

Microsoft Entra SSO enables single sign-on to the data gateway to access cloud data
sources that rely on Microsoft Entra ID based authentication. When you configure
Microsoft Entra SSO on the on-premises data gateway for an applicable data source,
queries run under the Microsoft Entra identity of the user that interacts with the Power
BI report.

While Azure Virtual Networks (VNets) offer network isolation and security for your
resources on the Microsoft cloud, you now require a secure way to connect to these
data sources. On-premises data gateways help you achieve that. Microsoft Entra SSO as
explained previously allows users to see only data that they have access to.

７ Note

VNet data gateways, which are available in public preview for Power BI Premium
Semantic models, eliminate the need to install an on-premises data gateway for
connecting to your VNet data sources. To learn more about VNet gateways and
their current limitations, see What is a virtual network (VNet) data gateway.

The following data sources aren't supported with Microsoft Entra SSO using an on-
premises data gateway behind an Azure VNet:

Analysis Services
ADLS Gen1
ADLS Gen2
Azure Blobs
CDPA
Exchange
OData
SharePoint
SQL Server
Web
AzureDevOpsServer
CDSTOData
Cognite
CommonDataService
Databricks
EQuIS
 Kusto (when using the newer “DataExplorer” function)
VSTS
Workplace Analytics

For more information on SSO, and a list of supported data sources for Microsoft Entra
SSO, see Overview of single sign-on for on-premises data gateways in Power BI.

Query steps when running Microsoft Entra SSO

Enable Microsoft Entra SSO for Gateway

Since the Microsoft Entra token of the user is passed via the gateway, it's possible for an
admin of the gateway computer to obtain access to these tokens. To make sure a user
with malicious intent isn't able to intercept these tokens, the following safeguard
mechanisms are available:

A tenant-level setting in the Power BI admin portal allows only Power BI service
admins to enable this feature for a tenant. For more information, see Microsoft
Entra single sign-on for gateways.
As a Power BI service admin, you can also control who can install gateways in your
tenant. For more information, see Manage gateway installers.

The Microsoft Entra SSO feature is disabled by default for on-premises data gateways.
As a Power BI admin, you must enable the Microsoft Entra Single Sign-On (SSO) for
Gateway tenant setting in the Power BI Admin portal before data sources are enabled
for Microsoft Entra SSO on an on-premises data gateway.
Related content
Overview of single sign-on for on-premises data gateways in Power BI
Troubleshoot gateways - Power BI
Article • 11/01/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

This article discusses some common issues when you use the on-premises data gateway
with Power BI. If you encounter an issue that isn't listed here, you can use the Power BI
Community site. Or, you can create a support ticket .

Configuration

Error: Power BI service reported local gateway as

unreachable. Restart the gateway and try again.
At the end of configuration, the Power BI service is called again to validate the gateway.
The Power BI service doesn't report the gateway as live. Restarting the Windows service
might allow the communication to be successful. To get more information, you can
collect and review the logs as described in Collect logs from the on-premises data
gateway app.

Data sources

７ Note

Not all data sources have dedicated articles detailing their connection settings or
configuration. For many data sources and non-Microsoft connectors, connection
options might vary between Power BI Desktop, and Manage gateways > Data
source settings configurations in the Power BI service. In such cases, the default
settings provided are the currently supported scenarios for Power BI.
Error: Unable to Connect. Details: "Invalid connection
credentials"
Within Show details, the error message that was received from the data source is
displayed. For SQL Server, you see a message like the following:

Output

Login failed for user 'username'.

Verify that you have the correct username and password. Also, verify that those
credentials can successfully connect to the data source. Make sure the account that's
being used matches the authentication method.

Error: Unable to Connect. Details: "Cannot connect to the

database"
You were able to connect to the server but not to the database that was supplied. Verify
the name of the database and that the user credential has the proper permission to
access that database.

Within Show details, the error message that was received from the data source is
displayed. For SQL Server, you see something like the following:

Output

Cannot open database "AdventureWorks" requested by the login. The login

failed. Login failed for user 'username'.

Error: Unable to Connect. Details: "Unknown error in data

gateway"
This error might occur for different reasons. Be sure to validate that you can connect to
the data source from the machine that hosts the gateway. This situation could be the
result of the server not being accessible.

Within Show details, you can see an error code of DM_GWPipeline_UnknownError.

You can also look in Event Logs > Applications and Services Logs > On-premises data
gateway Service for more information.
Error: We encountered an error while trying to connect to
<server>. Details: "We reached the data gateway, but the
gateway can't access the on-premises data source."
You were unable to connect to the specified data source. Be sure to validate the
information provided for that data source.

Within Show details, you can see an error code of

DM_GWPipeline_Gateway_DataSourceAccessError.

If the underlying error message is similar to the following, this means that the account
you're using for the data source isn't a server admin for that Analysis Services instance.
For more information, see Grant server admin rights to an Analysis Services instance.

Output

The 'CONTOSO\account' value of the 'EffectiveUserName' XML for Analysis

property is not valid.

If the underlying error message is similar to the following, it could mean that the service
account for Analysis Services might be missing the token-groups-global-and-universal
(TGGAU) directory attribute.

Output

The username or password is incorrect.

Domains with pre-Windows 2000 compatibility access have the TGGAU attribute
enabled. Most newly created domains don't enable this attribute by default. For more
information, see Some applications and APIs require access to authorization information
on account objects .

To confirm whether the attribute is enabled, follow these steps.

1. Connect to the Analysis Services machine within SQL Server Management Studio.
Within the Advanced connection properties, include EffectiveUserName for the
user in question and see if this addition reproduces the error.

2. You can use the dsacls Active Directory tool to validate whether the attribute is
listed. This tool is found on a domain controller. You need to know what the
distinguished domain name is for the account and pass that name to the tool.

Console
 dsacls "CN=John Doe,CN=UserAccounts,DC=contoso,DC=com"

You want to see something similar to the following in the results:

Console

Allow BUILTIN\Windows Authorization Access Group

SPECIAL ACCESS for
tokenGroupsGlobalAndUniversal
READ PROPERTY

To correct this issue, you must enable TGGAU on the account used for the Analysis
Services Windows service.

Another possibility for "The username or password is incorrect."

This error could also be caused if the Analysis Services server is in a different domain
than the users and there isn't a two-way trust established.

Work with your domain administrators to verify the trust relationship between domains.

Unable to see the data gateway data sources in the Get Data
experience for Analysis Services from the Power BI service

Make sure that your account is listed in the Users tab of the data source within the
gateway configuration. If you don't have access to the gateway, check with the
administrator of the gateway and ask them to verify. Only accounts in the Users list can
see the data source listed in the Analysis Services list.

Error: You don't have any gateway installed or configured

for the data sources in this dataset.
Ensure that you've added one or more data sources to the gateway, as described in Add
a data source. If the gateway doesn't appear in the admin portal under Manage
gateways, clear your browser cache or sign out of the service and then sign back in.

Error: Your data source can't be refreshed because the

credentials are invalid. Please update your credentials and
try again.
You were able to connect and refresh the dataset, with no runtime errors for the
connection, yet in the Power BI service this error bar appears. When the user attempts to
update the credentials with known-good credentials, an error appears stating that the
credentials supplied were invalid.

This error can occur when the gateway attempts a test connection, even if the
credentials supplied are acceptable and the refresh operation is successful. This occurs
because when the gateway performs a connection test, it does not include any optional
parameters during the connection attempt, and some data connectors (such as
Snowflake, for example) require optional connection parameters in order to connect.

When your refresh is completing properly and you do not experience runtime errors,
you can ignore these test connection errors for data sources that requires optional
parameters.

Datasets

Error: There is not enough space for this row.

This error occurs if you have a single row greater than 4 MB in size. Determine what the
row is from your data source, and attempt to filter it out or reduce the size for that row.

Error: The server name provided doesn't match the server

name on the SQL Server SSL certificate.
This error can occur when the certificate common name is for the server's fully qualified
domain name (FQDN), but you supplied only the NetBIOS name for the server. This
situation causes a mismatch for the certificate. To resolve this issue, make the server
name within the gateway data source and the PBIX file use the FQDN of the server.

Error: You don't see the on-premises data gateway

present when you configure scheduled refresh.
A few different scenarios could be responsible for this error:

The server and database name don't match what was entered in Power BI Desktop
and the data source configured for the gateway. These names must be the same.
They aren't case sensitive.
Your account isn't listed in the Users tab of the data source within the gateway
configuration. You need to be added to that list by the administrator of the
 gateway.
Your Power BI Desktop file has multiple data sources within it, and not all of those
data sources are configured with the gateway. You need to have each data source
defined with the gateway for the gateway to show up within scheduled refresh.

Error: The received uncompressed data on the gateway

client has exceeded the limit.
The exact limitation is 10 GB of uncompressed data per table. If you're hitting this issue,
there are good options to optimize and avoid it. In particular, reduce the use of highly
constant, long string values and instead use a normalized key. Or, removing the column
if it's not in use helps.

Error:
DM_GWPipeline_Gateway_SpooledOperationMissing
A few different scenarios could be responsible for this error

Gateway process might have restarted when the dataset refresh was in progress.
The gateway machine is cloned where gateway is running. We should not clone
gateway machine.

Reports

Error: Report could not access the data source because

you do not have access to our data source via an on-
premises data gateway.
This error is usually caused by one of the following:

The data source information doesn't match what's in the underlying dataset. The
server and database name need to match between the data source defined for the
on-premises data gateway and what you supply within Power BI Desktop. If you
use an IP address in Power BI Desktop, the data source for the on-premises data
gateway needs to use an IP address as well.
There's no data source available on any gateway within your organization. You can
configure the data source on a new or existing on-premises data gateway.
Error: Data source access error. Please contact the
gateway administrator.
If this report makes use of a live Analysis Services connection, you could encounter an
issue with a value being passed to EffectiveUserName that either isn't valid or doesn't
have permissions on the Analysis Services machine. Typically, an authentication issue is
due to the fact that the value being passed for EffectiveUserName doesn't match a local
user principal name (UPN).

To confirm the effective username, follow these steps.

1. Find the effective username within the gateway logs.

2. After you have the value being passed, validate that it's correct. If it's your user,
you can use the following command from a command prompt to see the UPN. The
UPN looks like an email address.

Console

whoami /upn

Optionally, you can see what Power BI gets from Azure Active Directory.

1. Browse to https://developer.microsoft.com/graph/graph-explorer.

2. Select Sign in in the upper-right corner.

3. Run the following query. You see a rather large JSON response.

HTTP

https://graph.windows.net/me?api-version=1.5

4. Look for userPrincipalName.

If your Azure Active Directory UPN doesn't match your local Active Directory UPN, you
can use the Map user names feature to replace it with a valid value. Or, you can work
with either your Power BI admin or local Active Directory admin to get your UPN
changed.

Kerberos
If the underlying database server and on-premises data gateway aren't appropriately
configured for Kerberos constrained delegation, enable verbose logging on the
gateway. Then, investigate based on the errors or traces in the gateway’s log files as a
starting point for troubleshooting. To collect the gateway logs for viewing, see Collect
logs from the on-premises data gateway app.

ImpersonationLevel
The ImpersonationLevel is related to the SPN setup or the local policy setting.

[DataMovement.PipeLine.GatewayDataAccess] About to impersonate user

DOMAIN\User (IsAuthenticated: True, ImpersonationLevel: Identification)

Solution

Follow these steps to solve the issue.

1. Set up an SPN for the on-premises gateway.

2. Set up constrained delegation in your Active Directory.

FailedToImpersonateUserException: Failed to create

Windows identity for user userid
The FailedToImpersonateUserException happens if you're unable to impersonate on
behalf of another user. This error could also happen if the account you're trying to
impersonate is from another domain than the one the gateway service domain is on.
This is a limitation.

Solution

Verify that the configuration is correct as per the steps in the previous
"ImpersonationLevel" section.
Ensure that the user ID it's trying to impersonate is a valid Active Directory
account.

General error: 1033 error while you parse the protocol

You get the 1033 error when your external ID that's configured in SAP HANA doesn't
match the sign-in if the user is impersonated by using the UPN ([email protected]). In
the logs, you see "Original UPN '[email protected]' replaced with a new UPN
'[email protected]'" at the top of the error logs, as seen here:

[DM.GatewayCore] SingleSignOn Required. Original UPN '[email protected]'

replaced with new UPN '[email protected].'

Solution

SAP HANA requires the impersonated user to use the sAMAccountName attribute
in Active Directory (user alias). If this attribute isn't correct, you see the 1033 error.

In the logs, you see the sAMAccountName (alias) and not the UPN, which is the
alias followed by the domain ([email protected]).

XML

<setting name="ADUserNameReplacementProperty" serializeAs="String">

<value>sAMAccount</value>
</setting>
<setting name="ADServerPath" serializeAs="String">
<value />
</setting>
<setting name="CustomASDataSource" serializeAs="String">
<value />
</setting>
 <setting name="ADUserNameLookupProperty" serializeAs="String">
<value>AADEmail</value>

[SAP AG][LIBODBCHDB DLL][HDBODBC] Communication

link failure:-10709 Connection failed (RTE:[-1] Kerberos
error. Major: "Miscellaneous failure [851968]." Minor: "No
credentials are available in the security package."
You get the "-10709 Connection failed" error message if your delegation isn't configured
correctly in Active Directory.

Solution

Make sure that you have the SAP Hana server on the delegation tab in Active
Directory for the gateway service account.

Export logs for a support ticket

Gateway logs are required for troubleshooting and creating a support ticket. Use the
following steps for extracting these logs.

1. Identify the gateway cluster.

 If you're a dataset owner, first check the gateway cluster name associated with
your dataset. In the following image, IgniteGateway is the gateway cluster.

2. Check the gateway properties.

The gateway admin should then check the number of gateway members in the
cluster and if load balancing is enabled.

If load balancing is enabled, then step 3 should be repeated for all gateway
members. If it's not enabled, then exporting logs on the primary gateway is
sufficient.

3. Retrieve and export the gateway logs.

Next, the gateway admin, who is also the administrator of the gateway system,
should do the following steps:

a. Sign in to the gateway machine, and then launch the on-premises data gateway
app to sign in to the gateway.

b. Enable additional logging.

c. Optionally, you can enable the performance monitoring features and include
performance logs to provide additional details for troubleshooting.

d. Run the scenario for which you're trying to capture gateway logs.

e. Export the gateway logs.

Refresh history
When you use the gateway for a scheduled refresh, Refresh history can help you see
what errors occurred. It can also provide useful data if you need to create a support
request. You can view scheduled and on-demand refreshes. The following steps show
how you can get to the refresh history.

1. In the Power BI nav pane, in Datasets, select a dataset. Open the menu, and select
Schedule refresh.

2. In Settings for..., select Refresh history.

For more information about troubleshooting refresh scenarios, see Troubleshoot refresh
scenarios.
Next steps
Troubleshoot the on-premises data gateway
Configure proxy settings for the on-premises data gateway
Manage your data source - Analysis Services
Manage your data source - SAP HANA
Manage your data source - SQL Server
Manage your data source - Import/scheduled refresh

More questions? Try the Power BI Community .

Troubleshoot Power BI gateway
(personal mode)
Article • 05/28/2024

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

The following sections go through some common issues you might come across when
you use the Power BI on-premises data gateway (personal mode).

Update to the latest version

The current version of the gateway for personal use is the on-premises data gateway
(personal mode). To use that version, update your installation.

Many issues can surface when the gateway version is out of date. It's a good general
practice to make sure you're on the latest version. If the date of the last the gateway
update is a month or longer, consider installing the latest version of the gateway. Then
attempt to reproduce the issue.

Installation
Gateway (personal mode) operates on 64-bit versions: If your computer is a 32-bit
version, you can't install the gateway (personal mode). Your operating system has to be
a 64-bit version. Install a 64-bit version of Windows or install the gateway (personal
mode) on a 64-bit computer.

Operation timed out: This message is common if the computer, physical or virtual
machine, on which you’re installing the gateway (personal mode) has a single core
processor. Close any applications, turn off any nonessential processes, and try installing
again.

Data management gateway or Analysis Services connector can't be installed on the

same computer as gateway (personal mode): If you already have an Analysis
Services connector or a data management gateway installed, you must first uninstall the
connector or the gateway. Then, try installing the gateway (personal mode).

７ Note

If you encounter an issue during installation, the setup logs can provide
information to help you resolve the issue. For more information, see Setup logs.

Proxy configuration: You might see issues with configuring the gateway (personal
mode) if your environment needs the use of a proxy. To learn more about how to
configure proxy information, see Configure proxy settings for the on-premises data
gateway.

Schedule refresh
Error: The credential stored in the cloud is missing.

You might get this error in settings for a dataset if you have a scheduled refresh and
then you uninstalled and reinstalled the gateway (personal mode). When you uninstall a
gateway (personal mode), the data source credentials for a dataset that was configured
for refresh are removed from the Power BI service.

Solution: In the Power BI service, go to the refresh settings for a dataset. In Manage
Data Sources, for any data source with an error, select Edit credentials. Then sign in to
the data source again.

Error: The credentials provided for the dataset are invalid. Please update the
credentials through a refresh or in the Data Source Settings dialog to continue.

Solution: If you get a credentials message, it could mean:

The usernames and passwords that you used to sign in to data sources aren't up to
date. In the Power BI service, go to refresh settings for the dataset. In Manage
Data Sources, select Edit credentials to update the credentials for the data source.

Mashups between a cloud source and an on-premises source, in a single query, fail
to refresh in the gateway (personal mode) if one of the sources is using OAuth for
authentication. An example of this issue is a mashup between CRM Online and a
local SQL Server instance. The mashup fails because CRM Online requires OAuth.

This error is a known issue, and it's being looked at. To work around the problem,
have a separate query for the cloud source and the on-premises source. Then, use
a merge or append query to combine them.
Error: Unsupported data source.

Solution: If you get an unsupported data source message in Schedule Refresh settings,
it could mean:

The data source isn't currently supported for refresh in the Power BI service.
The Excel workbook doesn't contain a data model, only worksheet data. The Power
BI service currently only supports refresh if the uploaded Excel workbook contains
a data model. When you import data by using Power Query in Excel, choose the
Load option to load data to a data model. This option ensures that data is
imported into a data model.

Error: [Unable to combine data] <query part>/<…>/<…> is accessing data sources

that have privacy levels, which cannot be used together. Please rebuild this data
combination.

Solution: This error is because of the privacy-level restrictions and the types of data
sources you're using.

Error: Data source error: We cannot convert the value "[Table]" to type Table.

Solution: This error is because of the privacy-level restrictions and the types of data
sources you're using.

Error: There is not enough space for this row.

Solution: This error occurs if you have a single row greater than 4 MB in size. Find the
row from your data source and filter out the row or reduce the size for that row.

Data sources
Missing data provider: The gateway (personal mode) operates on 64-bit versions only.
It requires a 64-bit version of the data providers to be installed on the same computer
where the gateway (personal mode) is installed. For example, if the data source in the
dataset is Microsoft Access, you must install the 64-bit ACE provider on the same
computer where you installed the gateway (personal mode).

７ Note

If you have a 32-bit version of Excel, you can't install a 64-bit version ACE provider
on the same computer.
Windows authentication is not supported for Access database: The Power BI service
currently only supports Anonymous authentication for the Access database.

Error: Sign-in error when you enter credentials for a data source: If you get an error
like this one when you enter Windows credentials for a data source:

You might still be on an older version of the gateway (personal mode).

Solution: For more information, see Install the latest version of Power BI gateway
(personal mode) .

Error: Sign-in error when you select Windows authentication for a data source using
ACE OLEDB: If you get the following error when you enter data source credentials for a
data source using an ACE OLEDB provider:
The Power BI service doesn't currently support Windows authentication for a data
source using an ACE OLEDB provider.

Solution: To work around this error, select Anonymous authentication. For a legacy ACE
OLEDB provider, anonymous credentials are equal to Windows credentials.

Tile refresh
If you receive an error when dashboard tiles refresh, see Troubleshooting tile errors.

Tools for troubleshooting

Refresh history
With Refresh history, you can see what errors occurred and find useful data if you need
to create a support request. You can view both scheduled and on-demand refreshes.
Here's how you get to Refresh history.

1. In the Power BI service navigation pane, in Semantic models, select a dataset.

Open the More options (...) menu, and select Schedule refresh.
2. In Settings for..., select Refresh history.
Event logs
Several event logs can provide information. The first two, Data Management Gateway
and PowerBIGateway, are present if you're an admin on the machine. If you're not an
admin, and you're using the data gateway (personal mode), the log entries within the
Application log displays.
The Data Management Gateway and PowerBIGateway logs are present under
Application and Services Logs.

Fiddler trace
Fiddler is a free tool from Telerik that monitors HTTP traffic. You can see the
communication with the Power BI service from the client machine. This communication
might show errors and other related information.

Setup logs
If the gateway (personal mode) fails to install, a link to show the setup log displays. The
setup log can show you details about the failure. These logs are Windows Install logs,
also known as Microsoft Software Installer (MSI) logs. They can be fairly complex and
hard to read. Typically, the resulting error is at the bottom, but determining the cause of
the error isn't trivial. It could be a result of errors in a different log. It could also be a
result of an error higher up in the log.
Or, you can go to your Temp folder (%temp%) and look for files that start with
Power_BI_.

７ Note

Going to %temp% might take you to a subfolder of Temp. The Power_BI_ files are in
the root of the Temp directory. You might need to go up a level or two.

Related content
Configure proxy settings for the on-premises data gateway
Data refresh in Power BI
Use personal gateways in Power BI
 Troubleshooting tile errors
Troubleshoot gateways - Power BI

More questions? Try asking the Power BI Community .

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Use custom data connectors with an on-
premises data gateway
Article • 03/15/2023

７ Note

We've split the on-premises data gateway docs into content that's specific to
Power BI and general content that applies to all services that the gateway
supports. You're currently in the Power BI content. To provide feedback on this
article, or the overall gateway docs experience, scroll to the bottom of the article.

You use Power BI data connectors to connect to and access data from an application,
service, or data source. You can develop custom data connectors and use them in Power
BI Desktop.

If you build reports in Power BI Desktop that use custom data connectors, you can use
an on-premises data gateway to refresh those reports in the Power BI service.

To learn more about how to develop custom data connectors for Power BI, see the
DataConnectors SDK in GitHub. This site includes information on how to get started,
and samples for Power BI and Power Query.

Enable and use custom connectors

To enable using custom connectors, select Connectors in the on-premises data gateway
app. In Custom data connectors, under Load custom data connectors from folder,
browse to and select a folder that the user running the gateway service can access. The
default user is NT SERVICE\PBIEgwService. The gateway automatically loads the custom
connector files in that folder, and they appear in the list of data connectors.
 ７ Note

If you're using an on-premises data gateway (personal mode), you can upload your
Power BI report to the Power BI service and use the gateway to refresh it.

For an on-premises data gateway, you need to create a data source for your custom
connector. On the gateway settings page in the Power BI service, select the option to
enable using custom connectors with this cluster.
When you enable this option, you see your custom connectors as available data source
connections that you can add to this gateway cluster. After you create a data source that
uses your new custom connector, you can refresh Power BI reports by using the custom
connector in the Power BI service.
Considerations and limitations
Make sure the folder you create is accessible to the background gateway service.
Typically, folders under your users' Windows folders or system folders aren't
accessible. The on-premises data gateway app shows a message if the folder isn't
accessible. This limitation doesn't apply to the on-premises data gateway (personal
mode).

For custom connectors to work with the on-premises data gateway, they need to
implement a TestConnection section in the custom connector code. This section
isn't required when you use custom connectors with Power BI Desktop. For this
reason, you can have a connector that works with Power BI Desktop, but not with
the gateway. For more information on how to implement a TestConnection section,
see TestConnection.

If your custom connector is on a network drive, include the fully qualified path in
the on-premises data gateway app.

You can only use one custom connector data source when working in DirectQuery
mode. Multiple custom connector data sources don't work with DirectQuery.

Next steps
Manage your data source - Analysis Services
 Manage your data source - SAP HANA
Manage your data source - SQL Server
Manage your data source - Oracle
Manage your data source - Import/scheduled refresh
Configure proxy settings for the on-premises data gateway
Use Kerberos for single sign-on (SSO) from Power BI to on-premises data sources

More questions? Try asking the Power BI Community .

Troubleshoot Power BI Desktop startup
Article • 11/10/2023

This article describes and provides remedies for several circumstances where Power BI
can't open or can't connect to data sources.

Issues with opening encrypted PBIX files

You can't open encrypted PBIX files by using a Power BI Desktop version that doesn't
support information protection. If you need to continue using Power BI Desktop, update
to a version that supports information protection.

Solution: Select this link to directly download the latest Power BI Desktop installation
executable . The latest version of Power BI Desktop supports information protection
and can decrypt and open any encrypted PBIX file.

On-premises data gateway issues

Users who installed and are running earlier versions of the Power BI on-premises data
gateway can be blocked from opening Power BI Desktop. Previous versions of the on-
premises data gateway placed administrative policy restrictions on named pipes on the
local machine.

Solution: To resolve the issue associated with the on-premises data gateway and enable
Power BI Desktop to open, use one of the following options:

Install the latest version of the Power BI on-premises data gateway.

The latest version of the Power BI on-premises data gateway doesn't place named
pipe restrictions on the local machine, and allows Power BI Desktop to open
properly. If you need to continue using the Power BI on-premises data gateway,
the recommended resolution is to update it. Select this link to directly download
the latest Power BI on-premises data gateway installation executable .

Uninstall or stop the Power BI on-premises data gateway service. You can uninstall
the Power BI on-premises data gateway if you no longer need it. Or you can stop
the Power BI on-premises data gateway service, which removes the policy
restriction and allows Power BI Desktop to open.

Run Power BI Desktop with administrator privileges.

 You can launch Power BI Desktop as an administrator, which also allows Power BI
Desktop to successfully open. It's still recommended to install the latest version of
the Power BI on-premises data gateway.

Power BI Desktop is a multiprocess architecture, and several of these processes

communicate by using Windows named pipes. Other processes might interfere
with those named pipes. The most common reason for such interference is
security, including situations where antivirus software or firewalls block the pipes
or redirect traffic to a specific port.

Opening Power BI Desktop with administrator privilege might resolve that issue. If
you can't open Power BI Desktop with administrator privilege, ask your
administrator which security rules are preventing named pipes from properly
communicating. Then add Power BI Desktop and its subprocesses to the allowlists.

Issues connecting to SQL Server

When you attempt to connect to a SQL Server database, you might see a message
similar to the following error:

An error happened while reading data from the provider:

'Could not load file or assembly 'System.EnterpriseServices, Version=4.0.0.0,
Culture=neutral, PublicKeyToken=xxxxxxxxxxxxx' or one of its dependencies. Either a
required impersonation level was not provided, or the provided impersonation level is
invalid. (Exception from HRESULT: 0x80070542)'

Solution: You can often resolve the issue if you open Power BI Desktop as an
administrator before you make the SQL Server connection. Opening Power BI Desktop
as an administrator and establishing the connection registers the required DLLs. After
that, you no longer have to open Power BI Desktop as an administrator. However, if
you're connecting to SQL server with alternate Windows credentials, you have to open
Power BI Desktop as an administrator every time you connect.

"Unable to sign in" issues

You might see a message similar to the following error:

Unable to sign in. Sorry, we encountered an error while trying to sign you in. Details:
The underlying connection was closed: Could not establish trust relationship for the
SSL/TLS secure channel.
Solution: Disable the certification revocation check at Options and settings > Options >
Security > Certification Revocation. For details, see Certificate revocation check, Power
BI Desktop.

Issues starting the Microsoft Store version of

Power BI Desktop
You might see a message similar to the following error:

Hmmmm... can't reach this page. ms-pbi.pbi.microsoft.com's server IP address could

not be found. Application event log message - The description for Event ID 1 from
source

The message might include further information, such as the following details:

Either the component that raises this event is not installed on your local computer or
the installation is corrupted. You can install or repair the component on the local
computer.

Solution: Reinstall WebView2 by using the following steps, which don't require elevated
administrative permissions.

1. Uninstall webview2.
2. Reinstall webview2 by using this installation link .

Issues related to WebView2

Rarely, Power BI Desktop might fail to start and displays a gray window, or an error
message that mentions WebView2.
Most cases are caused by a program on your machine, usually antivirus software. To
verify whether a program is causing the issue, take the following steps:

1. Close Power BI Desktop.

2. Open Windows Settings > About > Advanced system settings and select
Environment Variables.

Select New under User variables and add variable name

WEBVIEW2_ADDITIONAL_BROWSER_ARGUMENTS with the value --disable-
features=RendererCodeIntegrity.

3. Start Power BI Desktop and verify that it starts successfully this time.

4. Delete the environment variable you set.

Solution: If the preceding steps fixed the issue, disable any software that might be
interfering with Power BI Desktop startup, or provide an exemption for the WebView2
process.

If you still have issues, submit a support incident to Power BI support , and provide the
following information:

WebView2 error reports. If you use the Microsoft Store version of Power BI
Desktop, the error reports are at c:\Users\<username>\Microsoft\Power BI Desktop
Store App\WebView2\EBWebView\Crashpad\reports or c:\Users\
<username>\Microsoft\Power BI Desktop Store
App\WebView2Elevated\EBWebView\Crashpad\reports.

If you use the downloaded .exe version of Power BI Desktop, the error reports are
at c:\Users\<username>\AppData\Local\Microsoft\Power BI
Desktop\WebView2\EBWebView\Crashpad\reports or c:\Users\
<username>\AppData\Local\Microsoft\Power BI
Desktop\WebView2Elevated\EBWebView\Crashpad\reports.

Your machine's Device ID, from Windows Settings > System > About.

Installer and update logs. Collect the following files from the following locations
by copying and pasting the paths into File Explorer and then copying the files to
another location. Some files have the same names, so be sure not to overwrite
them but instead rename them when copying.

Path File

%temp%\ msedge_installer.log

%ProgramData%\Microsoft\EdgeUpdate\Log\ MicrosoftEdgeUpdate.log

%windir%\Temp\ MicrosoftEdgeUpdate.log

%allusersprofile%\Microsoft\EdgeUpdate\Log\ MicrosoftEdgeUpdate.log

%systemroot%\Temp\ msedge_installer.log

%localappdata%\Temp\ msedge_installer.log

%localappdata%\Temp\ MicrosoftEdgeUpdate.log

Event Viewer logs. Start Event Viewer from the Start menu. In Event Viewer, go to
Applications and Services log > Microsoft > Windows > CodeIntegrity >
Operational. Right-click Operational in the left pane and choose Save All Events
As. Store the file somewhere you can retrieve it. Do the same for Windows Logs >
Application.

The ClientState key from Registry Editor. Open Registry Editor by searching for
regedit in Windows Search or the Start menu. In Registry Editor, navigate to
HKEY_LOCAL_MACHINE\SOFTWARE\WOW6432Node\Microsoft\EdgeUpdate\Cli
entState. Right-click ClientState in the left pane, choose Export, and save the file.

Process traces. Follow these steps to collect process traces by using Process
Monitor:
1. Download Process Monitor, extract the downloaded file, and run
Procmon.exe.

2. Stop capturing by selecting the open-square Capture button.

3. Clear all traces by selecting the Clear garbage can button.

4. Start capturing by selecting the Capture button.

5. Launch Power BI Desktop and wait for the error to appear.

6. Stop the capture by selecting the Capture button.

7. Save the traces by choosing File > Save. In the Save to File dialog box, select
All events and Native Process Monitor Format (PML), provide a path for the
file, and then select OK.
 8. Share the traces with the support team on request.

Extra diagnostic information. Use the Windows Assessment and Deployment Kit
to collect extra information.

1. Download the Windows Assessment and Deployment Kit.

2. After downloading, start adksetup.exe, select Install the Windows Assessment

and Development Kit to this computer, and then select Next:

3. Continue the wizard. On the Select the features you want to install page,
select Windows Performance Toolkit, and then select Install:
4. Complete the installation, and then start Windows Performance Recorder.

5. Download the EdgeWebView2_General_EventsOnly.wprp file to your

machine and unpack it.

6. In Windows Performance Recorder, choose More options.

7. Choose Add Profiles to add the EdgeWebView2_General_EventsOnly.wprp

profile that you downloaded in the previous step.
8. Choose Start to start recording.

9. With the recording running, start Power BI Desktop and make sure the
startup issue occurs.
 10. When you're done, choose Save to stop the recording and save the results to
your computer.

11. Provide all information you collected to the support team on request.

Data connection time-outs

When you try to create a new connection or connect to an existing Power BI semantic
model, Power BI Desktop might time out without establishing the connection. The
connection spinner might continue to turn, but the connection never completes.

This situation can happen if your machine has a security product such as Digital
Guardian or other security products installed. In some cases, the installed security
product can interfere with outgoing network connection request calls, causing the
connection attempt to time out or fail.

Solution: Try disabling the security product, and then attempt the connection again. If
the connection succeeds after you disable the security product, you know that the
security product was probably the cause of the connection failure.

Other launch issues

The Power BI documentation team strives to cover as many Power BI Desktop issues as
possible. The team regularly looks at issues that might affect many customers, and
includes them in articles.

If your issue isn't related to an on-premises data gateway, or if the resolutions in this
article don't work, you can submit a support incident to Power BI support .

Whenever you experience issues with Power BI Desktop, it's helpful to turn on tracing
and gather log files. Log files can help isolate and identify the issue. To turn on tracing in
Power BI Desktop, choose File > Options and settings > Options, select Diagnostics,
and then select Enable tracing. Power BI Desktop must be running to set this option,
but it's helpful for any future issues associated with opening Power BI Desktop.

Next steps
Get Power BI Desktop
Troubleshoot refresh scenarios
Article • 04/15/2024

This article describes different scenarios you might encounter when refreshing data
within the Power BI service.

７ Note

If you encounter a scenario that's not listed in this article, and if it's causing issues,
you can ask for further assistance on the community site , or you can create a
support ticket .

You should always ensure that basic requirements for refresh are met and verified:

Verify the gateway version is up to date.

Verify the report has a gateway selected. If there's no gateway selected, the data
source might have changed or might be missing.

After you confirm the requirements are met, take a look through the following sections
for more troubleshooting.

Email notifications
If you're coming to this article from an email notification, and you no longer want to
receive emails about refresh issues, contact your Power BI admin. Ask them to remove
your email, or an email list you're subscribed to, from the appropriate semantic models
in Power BI. An admin uses the following area in the Power BI admin portal.
Refresh using Web connector doesn't work
properly
If you have a Web connector script that's using the Web.Page function, and you've
updated your semantic model or report after November 18, 2016, you must use a
gateway for refresh to work properly.

Unsupported data source for refresh

When you configure a semantic model, you might get an error indicating the semantic
model uses an unsupported data source for refresh. For details, see Troubleshooting
unsupported data source for refresh.

Dashboard doesn't reflect changes after refresh

Wait 10-15 minutes for a refresh to be reflected in the dashboard tiles. If it still doesn't
show up, repin the visualization to the dashboard.
GatewayNotReachable when setting credentials
You might encounter a GatewayNotReachable error when you try to set credentials for a
data source, which can be the result of an outdated gateway. Install the latest gateway
and try again.

Processing Error: The following system error

occurred: Type Mismatch
This error could be an issue with your M script within your Power BI Desktop file or Excel
workbook. It can also be due to an out-of-date Power BI Desktop version.

Tile refresh errors

For a list of errors you might encounter with dashboard tiles, and explanations, see
Troubleshooting tile errors.

Refresh fails when updating data from sources

that use Microsoft Entra ID OAuth
The Microsoft Entra ID (Microsoft Entra ID) OAuth token, used by many different data
sources, expires in approximately one hour. Sometimes that token expires before the
data has finished loading, since the Power BI service waits for up to two hours when
loading data. In that situation, the data loading process can fail with a credentials error.

Data sources that use Microsoft Entra ID OAuth include Microsoft Dynamics CRM
Online, SharePoint Online (SPO), and others. If you’re connecting to such data sources,
and get a credentials failure when loading data takes more than an hour, OAuth might
be the reason.

Microsoft is investigating a solution that allows the data loading process to refresh the
token and continue. However, if your Dynamics CRM Online or SharePoint Online
instance is so large that it runs over the two-hour data-load threshold, the Power BI
service might report a data load time-out. This data load time-out also applies to other
Microsoft Entra ID OAuth data sources.

For refresh to work properly when connecting to a SharePoint Online data source by
using Microsoft Entra ID OAuth, you must use the same account that you use to sign in
to the Power BI service.
If you want to connect to a data source from Power BI service by using OAuth2, the data
source must be in the same tenant as the Power BI service. Currently, multitenant
connection scenarios aren’t supported with OAuth2.

Uncompressed data limits for refresh

The maximum size for semantic models imported into the Power BI service is 1 GB.
These semantic models are heavily compressed to ensure high performance. In addition,
in shared capacity, the service places a limit of 10 GB on the amount of uncompressed
data that is processed during refresh. This limit accounts for the compression, and
therefore is higher than the 1-GB maximum semantic model size. Semantic models in
Power BI Premium aren't subject to these limits. If refresh in the Power BI service fails for
this reason, reduce the amount of data being imported to Power BI and try again.

Scheduled refresh time-out

Scheduled refresh for imported semantic models time out after two hours. This time-out
is increased to five hours for semantic models in Premium workspaces. If you encounter
this limit, consider reducing the size or complexity of your semantic model, or consider
refactoring the large semantic model into multiple smaller semantic models.

Scheduled refresh disabled

If a scheduled refresh fails four times in a row, Power BI disables the refresh. Address the
underlying problem, and then re-enable the scheduled refresh.

However, if the semantic model resides in a workspace under Embedded capacity, and
that capacity is switched off, the first attempt at refresh will fail (since the capacity is
switched off), and in this circumstance its scheduled refresh is immediately disabled.

Access to the resource is forbidden

This error can occur because of expired cached credentials. Clear your internet browser
cache, then sign in to Power BI and go to https://app.powerbi.com?
alwaysPromptForContentProviderCreds=true to force an update of your credentials.

Data refresh failure because of password

change or expired credentials
Data refresh can also fail due to expired cached credentials. Clear your internet browser
cache, then sign in to Power BI and go to https://app.powerbi.com?
alwaysPromptForContentProviderCreds=true , which forces an update of your credentials.

Refresh a column of the ANY type that contains

TRUE or FALSE results in unexpected values
When you create a report in Power BI Desktop that has an ANY data type column
containing TRUE or FALSE values, the values of that column can differ between Power BI
Desktop and the Power BI service after a refresh. In Power BI Desktop, the underlying
engine converts the boolean values to strings, retaining TRUE or FALSE values. In the
Power BI service, the underlying engine converts the values to objects, and then
converts the values to -1 or 0.

Visuals created in Power BI Desktop by using such columns might behave or appear as
designed prior to a refresh event, but might change (due to TRUE/FALSE being
converted to -1/0) after the refresh event.

Resolve the error: Container exited

unexpectedly with code 0x0000DEAD
If you get the Container exited unexpectedly with code 0x0000DEAD error, try to
disable the scheduled refresh and republish the semantic model.

Refresh operation throttled by Power BI

Premium
A Premium capacity might throttle data refresh operations when too many semantic
models are being processed concurrently. Throttling can occur in Power BI Premium
capacities. When a refresh operation is canceled, the following error messages are
logged into the refresh history:

You've exceeded the capacity limit for semantic model refreshes. Try again when fewer
semantic models are being processed.

If the error occurs frequently, use the schedule view to determine whether the
scheduled refresh events are properly spaced. To understand the maximum number of
concurrent refreshes allowed per SKU, review the Capacities and SKUs table.
To resolve this error, you can modify your refresh schedule to perform the refresh
operation when fewer semantic models are being processed. You can also increase the
time between refresh operations for all semantic models in your refresh schedule on the
affected Premium capacity. You can retry the operation if you're using custom XMLA
operations.

Capacity level limit exceeded.

This error indicates you have too many semantic models running refresh at the same
time, based on the capacity your organization has purchased. You can retry the refresh
operation, or reschedule the refresh time to address this error.

Node level limit exceeded.

This error indicates a system error in Power BI Premium based on semantic models
residing on a given physical node. You can retry the refresh operation, or reschedule the
refresh time to address this error.

Dataflows or datamart failures in Premium

workspaces
Some connectors aren't supported for dataflows and datamarts in Premium workspaces.
When using an unsupported connector, you may receive the following error:
Expression.Error: The import "<"connector name">" matches no exports. Did you miss a
module reference?

The following connectors aren't supported for dataflows and datamarts in Premium
workspaces:

Linkar
Actian
AmazonAthena
AmazonOpenSearchService
BIConnector
DataVirtuality
DenodoForPowerBI
Exasol
Foundry
Indexima
IRIS
JethroODBC
Kyligence
 MariaDB
MarkLogicODBC
OpenSearchProject
QubolePresto
SingleStoreODBC
StarburstPresto
TibcoTdv

The use of the previous list of connectors with dataflows or datamarts is only supported
workspaces that are not Premium.

There was a problem refreshing the dataflow,

the gateway version you are using is not
supported
This error occurs if the version of the on-premises data gateway being used to refresh
your dataflow (Gen1 or Gen2) is out of support. Currently Microsoft supports only the
last six versions of the on-premises data gateway. Update your gateway to the latest
version, or a supported version to resolve this issue. Use the update an on-premises
data gateway article for guidance on updating gateways.

Related content
Data refresh in Power BI
Troubleshoot the On-premises data gateway
Troubleshooting the Power BI Gateway - Personal

More questions? Try asking the Microsoft Power BI Community .

Feedback
Was this page helpful?  Yes  No

Provide product feedback | Ask the community

Troubleshoot tile errors
Article • 11/10/2023

This article lists and explains the common errors that can occur with tile refresh in Power
BI. If an error that's not listed causes you problems, you can ask for assistance on the
Power BI community site or file a support ticket .

Error list
The following list explains and offers solutions for common tile refresh errors.

Power BI encountered an unexpected error while loading the model. Please try
again later.

or

Couldn't retrieve the data model. Please contact the dashboard owner to make
sure the data sources and model exist and are accessible.

Power BI couldn't access your data because the data source wasn't reachable. This
issue can happen if the data source was removed, renamed, or moved, if the
source is offline, or if permissions have changed. Check that the source is still in
the specified location and you still have permission to access it. If that isn't the
problem, the source might be slow. Try again later during a time when the load on
the source is less. If it's an on-premises source, the data source owner might be
able to provide more information.

You don't have permission to view this tile or open the workbook.

Contact the dashboard owner to make sure the data sources and model exist and
are accessible for your account.

Power BI visuals have been disabled by your administrator.

Your Power BI administrator has disabled using Power BI visuals for your
organization or your security group. You can't use Power BI visuals from the
Microsoft marketplace or import private visuals from a file. You can use only the
pre-packed set of visuals.

Data shapes must contain at least one group or calculation that outputs data.
Please contact the dashboard owner.
There's no data to display because the query is empty. Try adding some fields from
the field list to the visual and repinning it.

Can't display the data because Power BI can't determine the relationship
between two or more fields.

You'e trying to use two or more fields from tables that aren't related. You need to
remove the unrelated fields from the visual and then create a relationship between
the tables. Once you create the relationship, you can add the fields back to the
visual. You can use Power BI Desktop or Power Pivot for Excel for this process. For
more information, see Create and manage relationships in Power BI Desktop.

The groups in the primary axis and the secondary axis overlap. Groups in the
primary axis can't have the same keys as groups in the secondary axis.

This issue is usually transient, and typically happens when you're moving groups
from rows to columns. The error should disappear when you finish moving all the
groups. If you still see the message, try switching fields between the rows,
columns, or axis legend, or removing fields from the visual.

This visual has exceeded the available resources. Try filtering to decrease the
amount of data displayed.

The visual has tried to query too much data for Power BI to complete the result
with available resources. Try filtering the visual to reduce the amount of data in the
result.

We are not able to identify the following fields: {0}. Please update the visual with
fields that exist in the semantic model.

The field was probably deleted or renamed. You can remove the broken field from
the visual, add a different field, and repin it.

Couldn't retrieve the data for this visual. Please try again later.

This issue is usually transient. If you try again later and still see this message,
contact support .

Tiles continue to show unfiltered data after you enable single-sign on (SSO).

This issue can happen if the underlying semantic model uses DirectQuery mode or
a Live Connection to Analysis Services through an on-premises data gateway. In
this issue, the tiles continue to show unfiltered data after you enable SSO for the
data source, until the next tile refresh. At the next tile refresh, Power BI uses SSO as
configured, and the tiles show the data filtered according to the user identity.
 To see the filtered data immediately, you can force a tile refresh. Select the Refresh
icon at the upper right of a Power BI dashboard.

As a semantic model owner, you can also increase the tile refresh frequency to 15
minutes to accelerate tile refresh. Select the gear icon in the upper right corner of
the Power BI service, and then select Settings. On the Semantic models tab,
expand Scheduled refresh, and under Automatic dashboard tile and metric
refresh, change Refresh frequency. Make sure you reset the configuration to the
original refresh frequency after Power BI does the next tile refresh.

７ Note

Automatic dashboard tile and metric refresh is available only for semantic
models in DirectQuery or Live Connection modes. Semantic models in Import
mode don't need a separate tile refresh because the tiles refresh automatically
during the next scheduled data refresh.

Support contact
If you're still having problems, contact support and ask them to investigate further.

Next steps
Troubleshoot the on-premises data gateway
Troubleshoot Power BI personal gateway
More questions? Try the Power BI community site.
Troubleshooting unsupported data
source for refresh
Article • 01/23/2024

You might see an error when trying to configure a semantic model for scheduled
refresh.

Output

You cannot schedule refresh for this semantic model because it gets data
from sources that currently don't support refresh.

This issue happens when the data source you used, within Power BI Desktop, isn't
supported for refresh. You need to find the data source that you're using and compare
that against the list of supported data sources at Refresh data in Power BI.

Find the data source

If you aren't sure what data source was used, you can find that using the following steps
within Power BI Desktop.

1. In Power BI Desktop, make sure you are on the Report pane.

2. Select Transform data from the ribbon bar.

3. Select Advanced Editor.

4. Make note of the provider listed for the source. In this example, the provider is
ActiveDirectory.

5. Compare the provider with the list of supported data sources found in Power BI
data sources.
 ７ Note

For refresh issues related to dynamic data sources, including data sources that
include hand-authored queries, see Refresh and dynamic data sources.

Related content
Data Refresh
Power BI Gateway - Personal
On-premises data gateway
Troubleshooting the On-premises data gateway
Troubleshooting the Power BI Gateway - Personal

More questions? Try asking the Power BI Community

Troubleshoot scheduled refresh for
Azure SQL databases in Power BI
Article • 02/20/2023

For detailed information about refresh, see Refresh data in Power BI and Configure
scheduled refresh.

While you set up scheduled refresh for an Azure SQL database, if you get an error with
error code 400 when editing the credentials, try these steps to configure the correct
firewall rule:

1. Sign in to the Azure portal .

2. Go to the Azure SQL database for which you're configuring refresh.

3. In the Overview page, select Set server firewall.

4. On the Networking page, select Allow Azure services and resources to access this
server and choose Save.

More questions? Ask the Power BI Community .

Troubleshoot the connection from Excel
to Power BI data
Article • 01/23/2024

There may be times when connecting Excel to Power BI data that you get an unexpected
result, or the feature doesn't work as you expected. This page provides solutions for
common issues when analyzing Power BI data in Excel.

７ Note

There are separate articles for different connection types. Those articles are as
follows:

Start in Power BI with Analyze in Excel.

Start in Excel to connect to Power BI semantic models.

If you encounter a scenario that's not listed below, ask for assistance on the Power
BI community site , or create a support ticket .

If you need to troubleshoot an issue with Power BI data in Excel, see the following
sections:

Forbidden error
Unable to access on-premises Analysis Services
Can't drag anything to the PivotTable Values area (no measures)

If you need to troubleshoot an issue in Power BI with Analyze in Excel, see the following
sections:

Connection cannot be made error

Can't find OLAP cube model error
Token expired error

Forbidden error
A user may have more than one Power BI account. When Excel tries to connect to Power
BI by using credentials from one of those accounts, it may attempt to use credentials
that don't have access to the desired semantic model or report.
When this situation occurs, you may receive an error titled Forbidden. This error means
you may be signed into Power BI with credentials that don't have permission to access
the semantic model. After encountering the Forbidden error and when you see the
prompt, type the credentials that have permission to access the semantic model you're
trying to use.

If you still run into errors, log into Power BI with the account that has permission. Then,
verify that you can view and access the semantic model in Power BI that you're
attempting to access in Excel.

Unable to access on-premises Analysis Services

If you're trying to access a semantic model that has a live connection to SQL Server
Analysis Services or Azure Analysis Services data, you may receive an error message. This
error may occur because a user can't connect to Power BI semantic models. This
situation may happen when you build semantic models on live connections to Analysis
Services unless the user has read access to the data in Analysis Services in addition to
the semantic models permissions in Power BI.

Can't drag anything to the PivotTable Values

area
Excel connects to Power BI through an external OLAP model. When these applications
connect, the PivotTable requires you to define measures in the external model because
all calculations are performed on the server. This requirement is different from working
with a local data source, such as tables in Excel and working with semantic models in
Power BI Desktop or the Power BI service). In those cases, the tabular model is available
locally, and you can use implicit measures . Implicit measures are generated
dynamically, and not stored in the data model. In these cases, the behavior in Excel is
different from the behavior in Power BI Desktop or the Power BI service. For instance,
there may be columns in the data that can be treated as measures in Power BI, but can't
be used as measures, or values, in Excel.

To address this issue, you have a few options:

Create measures in your data model in Power BI Desktop. Then, publish the data
model to the Power BI service and access that published semantic model from
Excel.
Create measures in your data model from Excel PowerPivot .
 If you imported data from an Excel workbook that had only tables and no data
model, then you can add the tables to the data model . Then, follow the steps in
the previous step to create measures in your data model.

Once you define your measures in the model in the Power BI service, you can use them
in the Values area in Excel PivotTables.

Connection cannot be made

The primary cause for a Connection cannot be made error is that your computer's OLE
DB provider client libraries aren't current.

Can't find OLAP cube model

The primary cause for a Can't find OLAP cube model error is that the semantic model
you're trying to access has no data model, and therefore the semantic model can't be
analyzed in Excel.

Token expired error

The primary cause for a Token expired error is that you haven't recently used the
Analyze in Excel feature on the computer you're using. To resolve this error, reenter
your credentials or reopen the file, and the error should go away.

Related content
Analyze in Excel

Tutorial: Create your own measures in Power BI Desktop

Measures in PowerPivot

Create a Measure in PowerPivot

Add worksheet data to a Data Model using a linked table

Error: We couldn't find any data in your
Excel workbook
Article • 11/10/2023

７ Note

This article applies to Excel 2007 and later.

When you import an Excel workbook into the Power BI service, you might see the
following error:

Output

Error: We couldn't find any data formatted as a table. To import from Excel
into the Power BI service, you need to format the data as a table. Select
all the data you want in the table and press Ctrl+T.

Quick solution
1. Edit your workbook in Excel.
2. Select the range of cells that contain your data. The first row should contain your
column headers, the column names.
3. Press Ctrl + T to create a table.
4. Save your workbook.
5. Return to the Power BI service and import your workbook again, or if you're
working in Excel 2016 and you've saved your workbook to OneDrive for work or
school, in Excel, select File > Publish.

Details

Cause
In Excel, you can create a table out of a range of cells, which makes it easier to sort,
filter, and format data.

When you import an Excel workbook, Power BI looks for these tables and imports them
into a semantic model. If it doesn't find any tables, you see this error message.

Solution
1. Open your workbook in Excel.

７ Note

The pictures here are of Excel 2013. If you're using a different version, things
might look a little different, but the steps are the same.

2. Select the range of cells that contain your data. The first row should contain your
column headers, the column names.
3. In the ribbon on the INSERT tab, select Table. Or, as a shortcut, press Ctrl + T.
4. You see the following dialog. Make sure My table has headers is selected, then
choose OK.

Now your data is formatted as a table.

5. Save your workbook.

6. Return to the Power BI service. Select Create, then choose these options.
 

7. In the Add data to get started window, select Excel.

8. Import your Excel workbook again. This time, the import should find the table and
succeed.

If the import still fails, let us know by selecting Community in the help menu:
Troubleshoot Access and Excel XLS
import issues in Power BI Desktop
Article • 03/12/2024

In Power BI Desktop, imported Access databases and Excel 97-2003 XLS files both use
the Access Database Engine. Three common situations can prevent the Access Database
Engine from working properly:

No Access Database Engine is installed.

The Access Database Engine bit version, 32-bit or 64-bit, is different from the
Power BI Desktop bit version.
You're using Access or XLS files with a Microsoft 365 subscription.

No Access Database Engine installed

If a Power BI Desktop error message indicates the Access Database Engine isn't installed,
install the Access Database Engine from the downloads page. Install the version,
either 32-bit or 64-bit, that matches your Power BI Desktop version.

If you work with dataflows and use a gateway to connect to the data, you must install
the Access Database Engine on the computer that runs the gateway.

７ Note

If the Access Database Engine bit version you install is different from your Microsoft
Office bit version, your Office applications won't be able to use the Access
Database Engine.

Access Database Engine bit version is different

from Power BI Desktop bit version
This situation usually occurs when the installed Microsoft Office version is 32-bit and the
installed Power BI Desktop version is 64-bit. The opposite can also happen, and the bit
version mismatch occurs in either case.

Any of the following solutions can remedy this bit-version mismatch error. You can also
apply these solutions to other mismatches, for example other 32-bit COM applications
like Visual Studio SSDT.
If you're using Access or XLS files with a Microsoft 365 subscription, see Access or XLS
files with Microsoft 365 for a different issue and resolution.

Solution 1: Change Power BI Desktop bit version to match

Microsoft Office bit version
To change the bit version of Power BI Desktop, uninstall Power BI Desktop, and then
install the version of Power BI Desktop that matches your Office installation.

７ Note

If you use the 32-bit version of Power BI Desktop to create very large data models,
you might experience out-of-memory issues.

To select a version of Power BI Desktop:

1. On the Power BI Desktop download page , choose your language, and then
select Download.

2. On the next screen, select the checkbox next to PBIDesktop.msi for the 32-bit
version, or PBIDesktop_x64.msi for the 64-bit version, and then select Next.

Solution 2: Change Microsoft Office bit version to match

Power BI Desktop bit version
To change the bit version of Microsoft Office to match the bit version of your Power BI
Desktop installation:

1. Uninstall Microsoft Office.

2. Install the version of Office that matches your Power BI Desktop installation.

Solution 3: Save the XLS file as XLSX

If the error occurs with an Excel 97-2003 XLS workbook, you can avoid using the Access
Database Engine by opening the XLS file in Excel and saving it as an XLSX file.

Solution 4: Install both versions of the Access Database

Engine
You can install both versions of the Access Database Engine to resolve the issue for
Power Query for Excel and Power BI Desktop. This workaround isn't recommended,
because it can introduce errors and issues for applications that use the Access Database
Engine bit version you installed first.

To use both Access Database Engine bit versions:

1. Install both bit versions of the Access Database Engine from the download page.

2. Run each version of the Access Database Engine by using the /passive switch. For
example:

Console

c:\users\joe\downloads\AccessDatabaseEngine.exe /passive

c:\users\joe\downloads\AccessDatabaseEngine_x64.exe /passive

You use Access or XLS files with Microsoft 365

Office 2013 and Office 2016 Microsoft 365 subscriptions register the Access Database
Engine provider in a virtual registry location that only Microsoft Office processes can
access. The Mashup Engine, which is responsible for running non-Microsoft 365 Excel
and Power BI Desktop, isn't an Office process, so it can't use the Access Database Engine
provider.
To fix this situation, download and install the Access Database Engine Redistributable
that matches the bit version of your Power BI Desktop installation, 32-bit or 64-bit.

Other import issues

The Power BI team regularly looks for issues that might affect many users, and tries to
include them in documentation. If you encounter an issue that this article doesn't cover,
submit a question about the issue to Power BI Support .
Troubleshoot DirectQuery models in
Power BI Desktop
Article • 02/28/2023

This article helps you diagnose performance issues with Power BI DirectQuery data
models you develop in Power BI Desktop or the Power BI service. The article also
describes how to get detailed information to help you optimize reports.

You should start any diagnosis of performance issues in Power BI Desktop, rather than in
the Power BI service or Power BI Report Server. Performance issues often depend on the
performance level of the underlying data source. You can more easily identify and
diagnose these issues in the isolated Power BI Desktop environment, without involving
components like an on-premises gateway.

If you don't find the performance issues in Power BI Desktop, you can focus your
investigation on the specifics of the report in the Power BI service.

You should also try to isolate issues to an individual visual before you look at many
visuals on a page.

Performance Analyzer
Performance Analyzer is a useful tool for identifying performance issues throughout the
troubleshooting process. If you can identify a single sluggish visual on a page in Power
BI Desktop, you can use Performance Analyzer to determine what queries Power BI
Desktop sends to the underlying source.

You also might be able to view traces and diagnostic information that the underlying
data sources emit. Such traces can contain useful information about the details of how
the query executed, and how to improve it.

Even without traces from the source, you can view the queries Power BI sent, along with
their execution times.

７ Note

For DirectQuery SQL-based sources, Performance Analyzer shows queries only for
SQL Server, Oracle, and Teradata data sources.
Trace file
By default, Power BI Desktop logs events during a given session to a trace file called
FlightRecorderCurrent.trc. You can find the trace file for the current session in the
AppData folder for the current user, at <User>\AppData\Local\Microsoft\Power BI
Desktop\AnalysisServicesWorkspaces.

The following DirectQuery data sources write all the queries that Power BI sends them to
the trace file. The log might support other DirectQuery sources in the future.

SQL Server
Azure SQL Database
Azure Synapse Analytics (formerly SQL Data Warehouse)
Oracle
Teradata
SAP HANA

To easily get to the trace file folder in Power BI Desktop, select File > Options and
settings > Options, and then select Diagnostics.
Under Crash Dump Collection, select the Open crash dump/traces folder link to open
the <User>\AppData\Local\Microsoft\Power BI Desktop\Traces folder.

Navigate to that folder's parent folder, and then open the AnalysisServicesWorkspaces
folder, which contains one workspace subfolder for every open instance of Power BI
Desktop. The subfolder names have integer suffixes, such as
AnalysisServicesWorkspace2058279583.

Each AnalysisServicesWorkspace folder includes a Data subfolder that contains the trace
file FlightRecorderCurrent.trc for the current Power BI session. This folder disappears
when the associated Power BI Desktop session ends.

You can open the trace files by using the SQL Server Profiler tool, which you can get as
part of the free SQL Server Management Studio (SSMS) download. After you download
and install SQL Server Management Studio, open SQL Server Profiler.
 

To open a trace file:

1. In SQL Server Profiler, select File > Open > Trace File.

2. Navigate to or enter the path to the trace file for the current Power BI session, such
as <User>\AppData\Local\Microsoft\Power BI
Desktop\AnalysisServicesWorkspaces\AnalysisServicesWorkspace2058279583\Data,
and open FlightRecorderCurrent.trc.

SQL Server Profiler displays all events from the current session. The following screenshot
highlights a group of events for a query. Each query group has the following events:

A Query Begin and Query End event, which represent the start and end of a DAX
query generated by changing a visual or filter in the Power BI UI, or from filtering
or transforming data in the Power Query Editor.

One or more pairs of DirectQuery Begin and DirectQuery End events, which
represent queries sent to the underlying data source as part of evaluating the DAX
query.
 

Multiple DAX queries can run in parallel, so events from different groups can interleave.
You can use the value of the ActivityID to determine which events belong to the same
group.

The following columns are also of interest:

TextData: The textual detail of the event. For Query Begin and Query End events,
the detail is the DAX query. For DirectQuery Begin and DirectQuery End events,
the detail is the SQL query sent to the underlying source. The TextData value for
the currently selected event also appears in the pane at the bottom of the screen.
EndTime: The time when the event completed.
Duration: The duration, in milliseconds, it took to run the DAX or SQL query.
Error: Whether an error occurred, in which case the event also displays in red.

The preceding image narrows some of the less interesting columns, so you can see the
more interesting columns more easily.

Follow this approach to capture a trace to help diagnose a potential performance issue:

1. Open a single Power BI Desktop session, to avoid the confusion of multiple

workspace folders.

2. Do the set of actions of interest in Power BI Desktop. Include a few more actions,
to ensure that the events of interest flush into the trace file.

3. Open SQL Server Profiler and examine the trace. Remember that closing Power BI
Desktop deletes the trace file. Also, further actions in Power BI Desktop don't
immediately appear. You must close and reopen the trace file to see new events.
Keep individual sessions reasonably small, perhaps 10 seconds of actions, not hundreds.
This approach makes it easier to interpret the trace file. There's also a limit on the size of
the trace file, so for long sessions, there's a chance of early events dropping.

Query and subquery format

The general format of Power BI Desktop queries is to use subqueries for each model
table the queries reference. The Power Query Editor query defines the subselect queries.
For example, assume you have the following TPC-DS tables in a SQL Server relational
database:

In the Power BI visual, the following expression defines the SalesAmount measure:

DAX

SalesAmount = SUMX(Web_Sales, [ws_sales_price] * [ws_quantity])

Refreshing the visual produces the T-SQL query in the following image. There are three
subqueries for the Web_Sales , Item , and Date_dim model tables. Each query returns all
the model table columns, even though the visual references only four columns.

These shaded subqueries are the exact definition of the Power Query queries. This use of
subqueries doesn't affect performance for the data sources DirectQuery supports. Data
sources like SQL Server optimize away the references to the other columns.

One reason Power BI uses this pattern is so you can define a Power Query query to use a
specific query statement. Power BI uses the query as provided, without an attempt to
rewrite it. This pattern restricts using query statements that use Common Table
Expressions (CTEs) and stored procedures. You can't use these statements in subqueries.
Gateway performance
For information about troubleshooting gateway performance, see Troubleshoot
gateways - Power BI.
Next steps
For more information about DirectQuery, check out the following resources:

Use DirectQuery in Power BI Desktop

Data sources supported by DirectQuery
DirectQuery models in Power BI Desktop
DirectQuery model guidance in Power BI Desktop

Questions? Try asking the Power BI Community

Troubleshooting nested values returned
as text in the Power BI service
Article • 02/10/2023

In the past, there have been cases where a report refreshes in Power BI Desktop, but fails
on the Power BI service with an error like this text:

Output

We cannot convert the value "[Table]" to type Table

Cause
One of the causes of this error involves nested non-scalar values, such as tables, records,
lists, and functions. When the Data Privacy Firewall buffers a data source, nested non-
scalar values are converted to text values, such as "[Table]" or "[Record]" .

The Power BI service now supports the setting of privacy levels or turning off the Firewall
entirely. The errors can be avoided by configuring the data source privacy settings in
the Power BI service to be non-Private.

For more recent versions of Power BI, when the Firewall buffers a nested table, record, or
list, it doesn't silently convert non-scalar values to text. Instead, it shows an error:

Output

We cannot return a value of type Table in this context

Effect on Load/Refresh
This change motivated by Firewall buffering extends to Load/Refresh, as well. The
behavior of loading nested tables, records, and lists to the Power BI Model and the Excel
Data Model in Power Query for Excel has changed. Before, nested items were loaded as
text values, such as "[Table]" or "[Record]" . Now, they're treated as errors. A null
value is in the loaded table and error count increments in the load results.

Since these errors only occur during Load/Refresh, they don't appear in the Power Query
Editor.
Before
Load/Refresh with no errors
Loaded table contains "[Table]" , "[Record]" , and so forth.

After
Load/Refresh with errors
Loaded table contains null , instead of "[Table]" , "[Record]" , and so forth.

Resolution
Are you loading a column that contains non-scalar values, for example, tables, lists, or
records? If so, you should be able to eliminate the errors by removing the column.

If you can't remove the column, try to replicate the old behavior by adding a custom
column and using logic like the following sample:

Output

if [MyColumn] is table then "[Table]" else if [MyColumn] is record then "

[Record]"
else if [MyColumn] is list then "[List]" else if [MyColumn] is function
then "[Function]" else [MyColumn]

Does the issue reproduce in Power BI Desktop if you set all your data source privacy
settings to Private? If so, try to resolve the error by configuring their data source privacy
settings in the Power BI service to be non-Private.
Data types in Power BI Desktop
Article • 11/10/2023

This article describes data types that Power BI Desktop and Data Analysis Expressions
(DAX) support.

When Power BI loads data, it tries to convert the data types of source columns into data
types that support more efficient storage, calculations, and data visualization. For
example, if a column of values you import from Excel has no fractional values, Power BI
Desktop converts the data column to a Whole number data type, which is better suited
for storing integers.

This concept is important because some DAX functions have special data type
requirements. In many cases DAX implicitly converts data types, but in some cases it
doesn't. For instance, if a DAX function requires a Date data type, but the data type for
your column is Text, the DAX function won't work correctly. So it's important and useful
to use the correct data types for columns.

Determine and specify a column's data type

In Power BI Desktop, you can determine and specify a column's data type in the Power
Query Editor, in Data View, or in Report View:

In Power Query Editor, select the column and then select Data Type in the
Transform group of the ribbon.
 In Data View or Report View, select the column, and then select the dropdown
arrow next to Data type on the Column tools tab of the ribbon.

The Data Type dropdown selection in Power Query Editor has two data types not
present in Data View or Report View: Date/Time/Timezone and Duration. When you
load a column with these data types into the Power BI model, a Date/Time/Timezone
column converts into a Date/time data type, and a Duration column converts into a
Decimal number data type.

The Binary data type isn't supported outside of the Power Query Editor. In the Power
Query Editor, you can use the Binary data type when you load binary files if you convert
it to other data types before loading it into the Power BI model. The Binary selection
exists in the Data View and Report View menus for legacy reasons, but if you try to load
Binary columns into the Power BI model, you might run into errors.

Number types
Power BI Desktop supports three number types: Decimal number, Fixed decimal
number, and Whole number.

You can use the Tabular Object Model (TOM) Column DataType property to specify the
DataType Enums for number types. For more information about programmatically
modifying objects in Power BI, see Program Power BI semantic models with the Tabular
Object Model.
Decimal number
Decimal number is the most common number type, and can handle numbers with
fractional values and whole numbers. Decimal number represents 64-bit (eight-byte)
floating point numbers with negative values from -1.79E +308 through -2.23E -308,
positive values from 2.23E -308 through 1.79E +308, and 0. Numbers like 34, 34.01, and
34.000367063 are valid decimal numbers.

The highest precision that the Decimal number type can represent is 15 digits. The
decimal separator can occur anywhere in the number. This type corresponds to how
Excel stores its numbers, and TOM specifies this type as DataType.Double Enum.

Fixed decimal number

The Fixed decimal number data type has a fixed location for the decimal separator. The
decimal separator always has four digits to its right, and allows for 19 digits of
significance. The largest value the Fixed decimal number can represent is positive or
negative 922,337,203,685,477.5807.

The Fixed decimal number type is useful in cases where rounding might introduce
errors. Numbers that have small fractional values can sometimes accumulate and force a
number to be slightly inaccurate. The Fixed decimal number type can help you avoid
these kinds of errors by truncating the values past the four digits to the right of decimal
separator.

This data type corresponds to SQL Server’s Decimal (19,4), or the Currency data type in
Analysis Services and Power Pivot in Excel. TOM specifies this type as DataType.Decimal
Enum.

Whole number
Whole number represents a 64-bit (eight-byte) integer value. Because it's an integer,
Whole number has no digits to the right of the decimal place. This type allows for 19
digits of positive or negative whole numbers between -9,223,372,036,854,775,807
(-2^63+1) and 9,223,372,036,854,775,806 (2^63-2), so can represent the largest
possible numbers of the numeric data types.

As with the Fixed decimal type, the Whole number type can be useful when you need
to control rounding. TOM represents the Whole number data type as DataType.Int64
Enum.
 ７ Note

The Power BI Desktop data model supports 64-bit integer values, but due to
JavaScript limitations, the largest number Power BI visuals can safely express is
9,007,199,254,740,991 (2^53-1). If your data model has larger numbers, you can
reduce their size through calculations before you add them to visuals.

Accuracy of number type calculations

Column values of Decimal number data type are stored as approximate data types,
according to the IEEE 754 Standard for floating point numbers. Approximate data types
have inherent precision limitations, because instead of storing exact number values, they
may store extremely close, or rounded, approximations.

Precision loss, or imprecision, can occur if the floating-point value can't reliably quantify
the number of floating point digits. Imprecision can potentially appear as unexpected or
inaccurate calculation results in some reporting scenarios.

Equality-related comparison calculations between values of Decimal number data type

can potentially return unexpected results. Equality comparisons include equals = ,
greater than > , less than < , greater than or equal to >= , and less than or equal to <= .

This issue is most apparent when you use the RANKX function in a DAX expression,
which calculates the result twice, resulting in slightly different numbers. Report users
might not notice the difference between the two numbers, but the rank result can be
noticeably inaccurate. To avoid unexpected results, you can change the column data
type from Decimal number to either Fixed decimal number or Whole number, or do a
forced rounding by using ROUND. The Fixed decimal number data type has greater
precision, because the decimal separator always has four digits to its right.

Rarely, calculations that sum the values of a column of Decimal number data type can
return unexpected results. This result is most likely with columns that have large
amounts of both positive numbers and negative numbers. The sum result is affected by
the distribution of values across rows in the column.

If a required calculation sums most of the positive numbers before summing most of
the negative numbers, the large positive partial sum at the beginning can potentially
skew the results. If the calculation happens to add balanced positive and negative
numbers, the query retains more precision, and therefore returns more accurate results.
To avoid unexpected results, you can change the column data type from Decimal
number to Fixed decimal number or Whole number.
Date/time types
Power BI Desktop supports five Date/Time data types in Power Query Editor. Both
Date/Time/Timezone and Duration convert during load into the Power BI Desktop data
model. The model supports Date/Time, or you can format the values as Date or Time
independently.

Date/Time represents both a date and time value. The underlying Date/Time value
is stored as a Decimal number type, so you can convert between the two types.
The time portion stores as a fraction to whole multiples of 1/300 seconds (3.33
ms). The data type supports dates between years 1900 and 9999.

Date represents just a date with no time portion. A Date converts into the model
as a Date/Time value with zero for the fractional value.

Time represents just a time with no date portion. A Time converts into the model
as a Date/Time value with no digits to the left of the decimal point.

Date/Time/Timezone represents a UTC date/time with a timezone offset, and

converts into Date/Time when loaded into the model. The Power BI model doesn't
adjust the timezone based on a user's location or locale. A value of 09:00 loaded
into the model in the USA displays as 09:00 wherever the report is opened or
viewed.

Duration represents a length of time, and converts into a Decimal Number type
when loaded into the model. As Decimal Number type, you can add or subtract
the values from Date/Time values with correct results, and easily use the values in
visualizations that show magnitude.

Text type
The Text data type is a Unicode character data string, which can be letters, numbers, or
dates represented in a text format. The practical maximum limit for string length is
approximately 32,000 Unicode characters, based on Power BI's underlying Power Query
engine, and its limits on text data type lengths. Text data types beyond the practical
maximum limit are likely to result in errors.

The way Power BI stores text data can cause the data to display differently in certain
situations. The next sections describe common situations that can cause Text data to
change appearance slightly between querying data in Power Query Editor and loading it
into Power BI.
Case sensitivity
The engine that stores and queries data in Power BI is case insensitive, and treats
different capitalization of letters as the same value. "A" is equal to "a". However, Power
Query is case sensitive, where "A" isn't the same as "a". The difference in case sensitivity
can lead to situations where text data changes capitalization seemingly inexplicably after
loading into Power BI.

The following example shows order data: An OrderNo column that's unique for each
order, and an Addressee column that shows the addressee name entered manually at
order time. Power Query Editor shows several orders with the same Addressee names
entered into the system with varying capitalizations.

After Power BI loads the data, capitalization of the duplicate names in the Data tab
changes from the original entry into one of the capitalization variants.

This change happens because Power Query Editor is case sensitive, so it shows the data
exactly as stored in the source system. The engine that stores data in Power BI is case
insensitive, so treats the lowercase and uppercase versions of a character as identical.
Power Query data loaded into the Power BI engine can change accordingly.

The Power BI engine evaluates each row individually when it loads data, starting from
the top. For each text column, such as Addressee, the engine stores a dictionary of
unique values, to improve performance through data compression. The engine sees the
first three values in the Addressee column as unique and stores them in the dictionary.
After that, because the engine is case insensitive, it evaluates the names as identical.
The engine sees the name "Taina Hasu" as identical to "TAINA HASU" and "Taina HASU",
so it doesn't store those variations, but refers to the first variation it stored. The name
"MURALI DAS" appears in uppercase letters, because that's how the name appeared the
first time the engine evaluated it when loading the data from top to bottom.

This image illustrates the evaluation process:

In the preceding example, the Power BI engine loads the first row of data, creates the
Addressee dictionary, and adds Taina Hasu to it. The engine also adds a reference to
that value in the Addressee column on the table it loads. The engine does the same for
the second and third rows, because these names aren't equivalent to the others when
ignoring case.

For the fourth row, the engine compares the value against the names in the dictionary
and finds the name. Since the engine is case insensitive, "TAINA HASU" and "Taina Hasu"
are the same. The engine doesn't add a new name to the dictionary, but refers to the
existing name. The same process happens for the remaining rows.

７ Note

Because the engine that stores and queries data in Power BI is case insensitive, take
special care when you work in DirectQuery mode with a case-sensitive source.
Power BI assumes that the source has eliminated duplicate rows. Because Power BI
is case insensitive, it treats two values that differ only by case as duplicate, whereas
the source might not treat them as such. In such cases, the final result is undefined.

To avoid this situation, if you use DirectQuery mode with a case-sensitive data
source, normalize casing in the source query or in Power Query Editor.

Leading and trailing spaces

The Power BI engine automatically trims any trailing spaces that follow text data, but
doesn't remove leading spaces that precede the data. To avoid confusion, when you
work with data that contains leading or trailing spaces, you should use the Text.Trim
function to remove spaces at the beginning or end of the text. If you don't remove
leading spaces, a relationship might fail to create because of duplicate values, or visuals
might return unexpected results.

The following example shows data about customers: a Name column that contains the
name of the customer and an Index column that's unique for each entry. The names
appear within quotes for clarity. The customer name repeats four times, but each time
with different combinations of leading and trailing spaces. These variations can occur
with manual data entry over time.

Row Leading space Trailing space Name Index Text length

1 No No "Dylan Williams" 1 14

2 No Yes "Dylan Williams " 10 15

3 Yes No " Dylan Williams" 20 15

4 Yes Yes " Dylan Williams " 40 16

In Power Query Editor, the resulting data appears as follows.

When you go to the Data tab in Power BI after you load the data, the same table looks
like the following image, with the same number of rows as before.

However, a visual based on this data returns just two rows.

In the preceding image, the first row has a total value of 60 for the Index field, so the
first row in the visual represents the last two rows of the loaded data. The second row
with total Index value of 11 represents the first two rows. The difference in the number
of rows between the visual and the data table is caused by the engine automatically
removing or trimming trailing spaces, but not leading spaces. So the engine evaluates
the first and second rows, and the third and fourth rows, as identical, and the visual
returns these results.

This behavior can also cause error messages related to relationships, because duplicate
values are detected. For example, depending on the configuration of your relationships,
you might see an error similar to the following image:

In other situations, you might be unable to create a many-to-one or one-to-one

relationship because duplicate values are detected.

You can trace these errors back to leading or trailing spaces, and resolve them by using
Text.Trim, or Trim under Transform, to remove the spaces in Power Query Editor.

True/false type
The True/false data type is a Boolean value of either True or False. For the best and most
consistent results, when you load a column that contains Boolean true/false information
into Power BI, set the column type to True/False.
Power BI converts and displays data differently in certain situations. This section
describes common cases of converting Boolean values, and how to address conversions
that create unexpected results in Power BI.

In this example, you load data about whether your customers have signed up for your
newsletter. A value of TRUE indicates the customer has signed up for the newsletter, and
a value of FALSE indicates the customer hasn't signed up.

However, when you publish the report to the Power BI service, the newsletter signup
status column shows 0 and -1 instead of the expected values of TRUE or FALSE. The
following steps describe how this conversion occurs, and how to prevent it.

The simplified query for this table appears in the following image:

The data type of the Subscribed To Newsletter column is set to Any, and as a result,
Power BI loads the data into the model as Text.

When you add a simple visualization that shows the detailed information per customer,
the data appears in the visual as expected, both in Power BI Desktop and when
published to the Power BI service.

However, when you refresh the semantic model in the Power BI service, the Subscribed
To Newsletter column in the visuals displays values as -1 and 0, instead of displaying
them as TRUE or FALSE:

If you republish the report from Power BI Desktop, the Subscribed To Newsletter
column again shows TRUE or FALSE as you expect, but once a refresh occurs in the
Power BI service, the values again change to show -1 and 0.

The solution to prevent this situation is to set any Boolean columns to type True/False in
Power BI Desktop, and republish your report.

When you make the change, the visualization shows the values in the Subscribed To
Newsletter column slightly differently. Rather than the text being all capital letters as
entered in the table, only the first letter is capitalized. This change is one result of
changing the column's data type.
Once you change the data type, republish to the Power BI service, and a refresh occurs,
the report displays the values as True or False, as expected.

To summarize, when working with Boolean data in Power BI, make sure your columns
are set to the True/False data type in Power BI Desktop.

Blank type
Blank is a DAX data type that represents and replaces SQL nulls. You can create a blank
by using the BLANK function, and test for blanks by using the ISBLANK logical function.

Binary type
You can use the Binary data type to represent any data with a binary format. In the
Power Query Editor, you can use this data type when loading binary files if you convert it
to other data types before you load it into the Power BI model.

Binary columns aren't supported in the Power BI data model. The Binary selection exists
in the Data View and Report View menus for legacy reasons, but if you try to load binary
columns to the Power BI model, you might run into errors.

７ Note

If a binary column is in the output of the steps of a query, attempting to refresh the
data through a gateway can cause errors. It's recommended that you explicitly
 remove any binary columns as the last step in your queries.

Table type
DAX uses a table data type in many functions, such as aggregations and time
intelligence calculations. Some functions require a reference to a table. Other functions
return a table that you can then use as input to other functions.

In some functions that require a table as input, you can specify an expression that
evaluates to a table. Some functions require a reference to a base table. For information
about the requirements of specific functions, see the DAX Function Reference.

Implicit and explicit data type conversion

Each DAX function has specific requirements for the types of data to use as inputs and
outputs. For example, some functions require integers for some arguments and dates
for others. Other functions require text or tables.

If the data in the column you specify as an argument is incompatible with the data type
the function requires, DAX may return an error. However, wherever possible DAX
attempts to implicitly convert the data to the required data type.

For example:

If you type a date as a string, DAX parses the string and tries to cast it as one of
the Windows date and time formats.
You can add TRUE + 1 and get the result 2, because DAX implicitly converts TRUE
to the number 1, and does the operation 1+1.
If you add values in two columns with one value represented as text ("12") and the
other as a number (12), DAX implicitly converts the string to a number, and then
does the addition for a numeric result. The expression = "22" + 22 returns 44.
If you try to concatenate two numbers, DAX presents them as strings, and then
concatenates. The expression = 12 & 34 returns "1234".

Tables of implicit data conversions

The operator determines the type of conversion DAX performs by casting the values it
requires before doing the requested operation. The following tables list the operators,
and the conversion DAX does on each data type when it pairs with the data type in the
intersecting cell.
 ７ Note

These tables don't include Text data type. When a number is represented in a text
format, in some cases Power BI tries to determine the number type and represent
the data as a number.

Addition (+)

INTEGER CURRENCY REAL Date/time

INTEGER INTEGER CURRENCY REAL Date/time

CURRENCY CURRENCY CURRENCY REAL Date/time

REAL REAL REAL REAL Date/time

Date/time Date/time Date/time Date/time Date/time

For example, if an addition operation uses a real number in combination with currency
data, DAX converts both values to REAL and returns the result as REAL.

Subtraction (-)

In the following table, the row header is the minuend (left side) and the column header
is the subtrahend (right side).

INTEGER CURRENCY REAL Date/time

INTEGER INTEGER CURRENCY REAL REAL

CURRENCY CURRENCY CURRENCY REAL REAL

REAL REAL REAL REAL REAL

Date/time Date/time Date/time Date/time Date/time

For example, if a subtraction operation uses a date with any other data type, DAX
converts both values to dates, and the return value is also a date.

７ Note

Data models support the unary operator, - (negative), but this operator doesn't
change the data type of the operand.
Multiplication (*)

INTEGER CURRENCY REAL Date/time

INTEGER INTEGER CURRENCY REAL INTEGER

CURRENCY CURRENCY REAL CURRENCY CURRENCY

REAL REAL CURRENCY REAL REAL

For example, if a multiplication operation combines an integer with a real number, DAX
converts both numbers to real numbers, and the return value is also REAL.

Division (/)
In the following table, the row header is the numerator and the column header is the
denominator.

INTEGER CURRENCY REAL Date/time

INTEGER REAL CURRENCY REAL REAL

CURRENCY CURRENCY REAL CURRENCY REAL

REAL REAL REAL REAL REAL

Date/time REAL REAL REAL REAL

For example, if a division operation combines an integer with a currency value, DAX
converts both values to real numbers, and the result is also a real number.

Comparison operators
In comparison expressions, DAX considers Boolean values greater than string values,
and string values greater than numeric or date/time values. Numbers and date/time
values have the same rank.

DAX doesn't do any implicit conversions for Boolean or string values. BLANK or a blank
value is converted to 0, "", or False, depending on the data type of the other compared
value.

The following DAX expressions illustrate this behavior:

=IF(FALSE()>"true","Expression is true", "Expression is false") returns

"Expression is true".
 =IF("12">12,"Expression is true", "Expression is false") returns "Expression is

true".

=IF("12"=12,"Expression is true", "Expression is false") returns "Expression is

false".

DAX does implicit conversions for numeric or date/time types as the following table
describes:

Comparison INTEGER CURRENCY REAL Date/time

Operator

INTEGER INTEGER CURRENCY REAL REAL

CURRENCY CURRENCY CURRENCY REAL REAL

REAL REAL REAL REAL REAL

Date/time REAL REAL REAL Date/Time

Blanks, empty strings, and zero values

DAX represents a null, blank value, empty cell, or missing value by the same new value
type, a BLANK. You can also generate blanks by using the BLANK function, or test for
blanks by using the ISBLANK function.

How operations such as addition or concatenation handle blanks depends on the

individual function. The following table summarizes the differences between how DAX
and Microsoft Excel formulas handle blanks.

Expression DAX Excel

BLANK + BLANK BLANK 0 (zero)

BLANK + 5 5 5

BLANK * 5 BLANK 0 (zero)

5/BLANK Infinity Error

0/BLANK NaN Error

BLANK/BLANK BLANK Error

FALSE OR BLANK FALSE FALSE

FALSE AND BLANK FALSE FALSE

 Expression DAX Excel

TRUE OR BLANK TRUE TRUE

TRUE AND BLANK FALSE TRUE

BLANK OR BLANK BLANK Error

BLANK AND BLANK BLANK Error

Next steps
You can do all sorts of things with Power BI Desktop and data. For more information on
Power BI capabilities, see the following resources:

What is Power BI Desktop?

Query overview with Power BI Desktop
Data sources in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop
Connector extensibility in Power BI
Article • 01/23/2024

Power BI can connect to data by using existing connectors and generic data sources, like
ODBC, OData, OLE DB, Web, CSV, XML, and JSON. Developers can also enable new data
sources with custom data extensions called custom connectors. Microsoft certifies and
distributes some custom connectors as certified connectors.

To use non-certified custom connectors that you or another party develop, you must
adjust your Power BI Desktop security settings to allow extensions to load without
validation or warning. These extensions can ignore privacy levels and handle credentials,
including sending them over HTTP, so you should use this setting only if you completely
trust your custom connectors.

Another option is for the developer to sign the connector with a certificate, and provide
the information you need to use the connector without changing your security settings.
For more information, see Trusted third-party connectors.

Custom connectors
Non-certified custom connectors can range from small business-critical APIs to large
industry-specific services that Microsoft hasn't released a connector for. Many
connectors are distributed by vendors. If you need a connector for a specific industry or
business, contact the vendor.

To use a non-certified custom connector:

1. Put the connector .pq, .pqx, .m, or .mez file in your local [Documents]\Microsoft
Power BI Desktop\Custom Connectors folder. If the folder doesn't exist, create it.

2. To adjust the data extension security settings, in Power BI Desktop, select File >
Options and settings > Options > Security.

3. Under Data Extensions, select (Not Recommended) Allow any extension to load
without validation or warning.

4. Select OK, and then restart Power BI Desktop.

The default Power BI Desktop data extension security setting is (Recommended) Only
allow Microsoft certified and other trusted third-party extensions to load. With this
setting, if there are non-certified custom connectors on your system, the Uncertified
Connectors dialog box appears at Power BI Desktop startup, listing the connectors that
can't securely load.
To clear the error if you don't need to use the connectors in this session, select OK.

To prevent the error, either change your Data Extensions security setting, or remove the
uncertified connectors from your Custom Connectors folder.

） Important

You can use only one custom connector data source when you work in DirectQuery
mode. Multiple custom connector data sources won't work with DirectQuery.

Certified connectors
Microsoft certifies a limited subset of custom data extensions. While Microsoft
distributes these connectors, Microsoft isn't responsible for their performance or
continued function. The third-party developer who created the connector is responsible
for its maintenance and support.

In Power BI Desktop, certified third-party connectors appear in the list in the Get Data
dialog box, along with generic and common connectors. You don't need to adjust
security settings to use the certified connectors.

Related content
To get a custom connector certified, see Power Query Connector Certification.
Trusted third-party connectors
Article • 01/23/2023

In Power BI Desktop, we generally recommend keeping your Data extension security

level at the higher level, which prevents loading of code not certified by Microsoft.
However, there might be many cases in which you want to load specific connectors.
These connectors include ones you've written and ones provided to you by a consultant
or vendor outside the Microsoft certification path.

The developer of a given connector can sign it with a certificate and provide you with
the information you need to securely load it without lowering your security settings.

For more information about the security settings, see Connector extensibility in Power
BI.

Using the registry to trust third-party

connectors
Trusting third-party connectors in Power BI Desktop is done by listing the thumbprint of
the certificate you want to trust in a specified registry value. If this thumbprint matches
the thumbprint of the certificate on the connector you want to load, you can load it in
the Recommended security level of Power BI Desktop.

The registry path is HKEY_LOCAL_MACHINE\Software\Policies\Microsoft\Power BI Desktop .

Make sure the path exists, or create it. This location is chosen due to it being primarily
controlled by IT policy and requiring local machine administration access to edit.
 

Add a new value under the path specified in the previous image. The type should be
Multi-String Value: REG_MULTI_SZ . It should be called TrustedCertificateThumbprints .
 

Add the thumbprints of the certificates you want to trust. You can add multiple
certificates by using \0 as a delimiter, or in the Registry Editor, right-click key, then
select Modify to put each thumbprint on a new line. This example thumbprint is taken
from a self-signed certificate.
 

If you have the proper thumbprint from your developer, you should now be able to
securely trust connectors signed with the associated certificate.

How to sign connectors

If you have a connector you or a developer need to sign, see Handling Power Query
Connector Signing.
Manage DirectQuery connections to a
published semantic model
Article • 11/10/2023

By default, when you publish a semantic model to the Power BI service, you can make a
DirectQuery connection to it, assuming you have proper permissions. You can use this
connection to create new composite models on top of the semantic model.

In some situations, however, you need to discourage these connections from

happening. Discouraging these connections is especially important in the composite
models scenario, where you might want to prohibit creation of new composite models
on top of the semantic model (so-called chaining). By discouraging DirectQuery
connections to a semantic model, you're effectively ending the chain or stopping it from
forming in the first place.

７ Note

Power BI honors this setting and disables making DirectQuery connections to a

semantic model, but third-party tools might not. Third-party tools might allow
users to make DirectQuery connections to a semantic model even if you disabled it.

Use Power BI Desktop to discourage

DirectQuery connections to a semantic model
1. To discourage DirectQuery connections to a semantic model, go to File > Options
and settings > Options > Current File > Published semantic model settings.

2. On this page, choose the Discourage DirectQuery connections option, and select
OK.
Use third-party tools to discourage
DirectQuery connections to a semantic model
By using third-party tools, you can discourage DirectQuery connections to a semantic
model by setting the DiscourageCompositeModels property on a model to True .

Next steps
Using DirectQuery in Power BI
Semantic models in the Power BI service
Use composite models in Power BI Desktop
More questions? Ask the Power BI Community
Troubleshooting sign-in with OData
feed
Article • 01/31/2023

This article contains troubleshooting options when signing in using an organization

account for an OData feed:

Credential type not supported error

Access denied errors

The following sections describe each error, and the steps to remedy them, in turn.

Credential type not supported

You might see the following error, indicating the credential type isn't supported:

Output

We are unable to connect because this credential type is not supported

by this resource. Please choose another credential type.

You need to ensure your service is sending auth headers as follows:

First Oauth request without any authorization header should send the following
header in response:

Output

www-authenticate: Bearer realm=https://login.microsoftonline.com/<Your

Active Directory Tenant Id>

Redirect request to the service with authorization header set to Bearer should send
the following header in response:

Output

www-authenticate: Bearer
authorization_uri=https://login.microsoftonline.com/<Your Active
Directory Tenant Id>/oauth2/authorize

After a successful redirect call, calls to your service have the right access token in the
authorization header. If you still see an error, clear the Global Permissions for the OData
service URI and try again. To clear Global Permissions, go to File > Options and
Settings > Data Source Settings > Global Permissions.

Access denied
You might see one of the following errors, indicating access is denied:

Output

access_denied: AADSTS650053: The application 'Microsoft Power Query for

Excel'
asked for scope 'user_impersonation' that doesn't exist
on the resource \<resourceId\>.

Output

Microsoft Power Query for Excel needs permission to access resources

in your organization that only an admin can grant.
Ask an admin to grant permission to this app before you can use it.

If you encounter such an error, ensure the application registration for your OData
service has following settings:

Application ID is set to the OData service base URI.

Scope user_impersonation is defined.
The application's permissions are appropriately set by the administrator.

Next steps
You can do all sorts of things with Power BI Desktop. For more information on its
capabilities, check out the following resources:

What is Power BI Desktop?

Query overview with Power BI Desktop
Data types in Power BI Desktop
Shape and combine data with Power BI Desktop
Common query tasks in Power BI Desktop