Best Practice - Data Consistency Check For Logistics PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 68
At a glance
Powered by AI
The document discusses best practices for checking and ensuring data consistency across different areas of an SAP system including Sales & Distribution, Materials Management, Supply Chain Management, and between different SAP modules. It provides definitions of various types of inconsistencies and their typical root causes. It also outlines general procedures for detecting and resolving inconsistencies as well as specific tools and reports available within SAP for this purpose.

The document discusses tools for checking consistency within SAP ECC, between ECC and CRM, and between ECC and SCM. It outlines reports and tools available in SAP CRM, SCM, and across different modules like SD, LE, MM and FI.

The document mentions that typical root causes of inconsistencies can include incorrect configuration, errors during data transfer between systems, incorrect master data, manual errors during data entry, and technical problems with interfaces or batch input.

Data Consistency Check for Logistics

Best Practice for Solution Management

Version Date: October 2008

Contents
1. Introduction .............................................................................................................................3
1.1 Applicability, Goals, and Usage of this Document.................................................................3
1.1.1 Goal of this Best Practice Document and Assessment Guide....................................3
1.1.2 How to use this Best Practice ...................................................................................3
1.1.3 Legend.....................................................................................................................4
2. General Best Practice Procedure for Inconsistencies...............................................................5
2.1 Background Information.......................................................................................................5
2.1.1 Definition of Terms and Types of Inconsistencies ......................................................5
2.1.2 Typical Root Causes for Inconsistencies...................................................................6
2.2 Overall Procedure................................................................................................................9
2.3 Assessment Guide.............................................................................................................13
2.3.1 Understanding the Situation and Inconsistency.......................................................13
2.3.2 Process Understanding ..........................................................................................15
3. Tools and Procedures within an SAP Environment.................................................................18
3.1 Tools to Check Consistency Between ECC/CRM and Within CRM .....................................18
3.1.1 Data Exchange for the Most Common CRM Scenario - CRM Sales........................18
3.1.2 Data Exchange for the CRM Scenario CRM Mobile ................................................19
3.1.3 DIMa – The Tool to Detect and Repair Inconsistencies Between SAP ECC and SAP
CRM 19
3.1.4 Analysis and Correction Reports in Data Exchange of One Order Documents
Between SAP CRM and R/3..................................................................................................20
3.1.5 Miscellaneous Check, Correction and Repair Reports in CRM ................................22
3.2 Tools to Check Consistency Between ECC/SCM and Within SCM......................................23
3.2.1 Introduction ............................................................................................................23
3.2.2 Product Allocation (PAL): Internal Inconsistencies Within SCM or Between ECC/SCM23
3.2.3 Time Series of DP and SNP ...................................................................................25
3.2.4 Shipping and Vehicle Scheduling............................................................................25
3.2.5 Integration Models (External Consistency)..............................................................25
3.2.6 Internal Consistency Check for SCM regarding LC <-> DB .....................................25
3.2.7 Consistency at Interface: Transaction Data (External Consistency) .........................26
3.2.8 Temporary Quantity Assignments (TQAs) ...............................................................27
2
Best Practice: Data Consistency Check
3.2.9 Transaction Data: All Order Types and Stock (External Consistency) ......................33
3.2.10 Sales order requirements and SD mapping tables (external consistency)................35
3.3 Tools to Check Consistency within ECC.............................................................................37
3.3.1 Tools for Processes Involving SD and LE ...............................................................37
3.3.2 Tools for Processes Involving MM...........................................................................47
3.3.3 Inconsistencies between MM and FI.......................................................................49
3.3.4 Tools for Processes Involving WM ..........................................................................51
3.3.5 Tools for Processes Involving PP............................................................................52
3.3.6 Tools for Processes Involving PS............................................................................53
3.3.7 Tools to Check the Logistics Information System ....................................................54
3.4 Tools not Depending on a Particular System ......................................................................55
3.4.1 Tools for ALE Consistency Check ...........................................................................55
3.4.2 Tools for ALE Recovery ..........................................................................................55
3.4.3 Generic Consistency Check for Two Linked Tables in One System .........................56
3.4.4 The Generic External Compare ..............................................................................60
4. Appendix...............................................................................................................................61
4.1 General Roadmap for Analysis...........................................................................................61
4.2 Dos and don’ts for Data Consistency .................................................................................64
4.3 Advantages and Disadvantages of Different Recovery Methods .........................................64
4.4 Further Information ............................................................................................................67
4.4.1 Related information ................................................................................................67
4.4.2 Troubleshooting .....................................................................................................67

2
© 2008 SAP AG
1. Introduction
1.1 Applicability, Goals, and Usage of this Document
To ensure that this Best Practice is the one you need, consider the following goals and
requirements.
1.1.1 Goal of this Best Practice Document and Assessment Guide
This Best Practice provides information how to monitor data consistency and what you
should do once an inconsistency is reported. The document provides the general monitoring
procedure as well as details for several areas where potentially business critical
inconsistencies could occur. The more detailed sections should facilitate the use of available
consistency check tools by providing the general description and logical design of such
reports including relevant business data needed for some frequent instances of
inconsistencies. The goal is to enable you to use them and to understand the guiding
principles behind these tools, so that they may be available as templates for specific cases
involving non-SAP systems.
The Best Practice document is divided into two major areas: Section 2: General Best
Practice Procedure for Inconsistencies provides background information about typical root
causes of inconsistencies, as well as the basic questions you need to ask to identify the most
likely root causes for the specific case you encounter. Section 3: Tools and Procedures
within an SAP Environment describes the most important tools available from SAP and how
they can be used to monitor data consistency on a regular basis and correct possible
inconsistencies. Section 4: Appendix provides additional information like analysis roadmaps
and a comparison of different recovery methods.
In addition to the consistency check methods, the Best Practice document provides basic
deciding factors for recovery methods to determine which is preferable in case of data loss
and how to evaluate them for a given case. Please check the Best Practice Document
“Business Continuity Management for SAP System Landscapes” for a more specific and
detailed view on the area of data recovery and business continuity.

1.1.2 How to use this Best Practice


Read through section 2: General Best Practice Procedure for Inconsistencies to get
background information about inconsistencies and common root causes for inconsistencies.
Afterwards, you should understand the general procedure so that you can deal with specific
new cases. The result of working through this section will be a flow chart for the business
processes where the difference has been reported, including an understanding of the logical
steps, as well as an understanding of the involved technical data.
Afterwards, you can set up your data consistency monitoring and investigations of specific
root causes using the information available in section 3: Tools and Procedures within an SAP
Environment. If you have enhanced your solution, you should consult section 3: Tools and
Procedures within an SAP Environment as well to check whether some reports could be
used as a template in your situation. You will find a clear link to the corresponding section in
the generic part.
Section 4: Appendix provides a graphical roadmap and deciding factors for typical data
recovery methods. You can use this section to get a quick overview as well as a basis to
discuss your own goals and procedures.
4
Best Practice: Data Consistency Check

1.1.3 Legend
This symbol indicates a paragraph from where you can navigate to another section of this
document for more detailed information on a topic.

This symbol indicates a paragraph from where you can navigate to another document within
the SAP Service Marketplace for more detailed information on a topic.

This symbol indicates an important paragraph where you find special hints or best practices.

4
© 2008 SAP AG
5
Best Practice: Data Consistency Check

2. General Best Practice Procedure for


Inconsistencies
2.1 Background Information
This part will build up the basic understanding needed to deal with inconsistencies and is the
base line for the more detailed discussions later on in this document.
2.1.1 Definition of Terms and Types of Inconsistencies
It is very important to differentiate between the terms Difference and Inconsistency. While
Difference relates to a temporary data mismatch that occurs due to processing time of
asynchronous processes, Inconsistency means a permanent mismatch that does not
disappear once all system activities are completed successfully.
Real inconsistencies can be classified by the different root causes and general behavior:
A Technical Inconsistency is everything that can be found on a database level and needs
appropriate correction in the underlying database, while a Logical Inconsistency is a
permanent mismatch that is due to a misunderstanding of process or misinterpretation of
data. While technical inconsistencies can be identified by technical means like check tools,
logical inconsistencies need to be identified by mapping the intended business process to the
underlying data structures. They cannot be identified by technical means, as the underlying
data is consistent on a one-to-one level.
Looking at a business object instance A.obj in a system A and the respective business object
instance B.obj in system B, the following inconsistencies for the business object instance
between both environments may occur:

Difference System A System B Difference type description


type
1 A.obj B.obj Some value in the business object instance of A
and B differs
2 A.obj Respective Business object instance is present in A but
instance is missing in B
missing
3 Respective B.obj Business object instance is present in B but
instance is missing in A
missing

Any good consistency check tool should compare current business object instances between
the two systems, A and B, display all differences categorized into the three difference types
and – if possible – provide drilldown functionality into the exact different fields for difference
type 1.
Three different cases are usually distinguished when investigating inconsistencies:
Inconsistencies between the real world and a system
Inconsistencies between two systems
Inconsistencies within one system

As the data needs to be compared between the two environments independently of the
nature of the environment, and the investigation needs to be based on tools for easy
handling the general procedure to identify the inconsistencies and the data correction will
always take place in a computer system (in case 1, the real world will always be the “leading
system”). Thus, a unified overall procedure can be applied and we will not distinguish
between the three cases unless it is essential for a specific procedure or understanding how
to apply a procedure in the given environment.

5
© 2008 SAP AG
6
Best Practice: Data Consistency Check

2.1.2 Typical Root Causes for Inconsistencies


Most inconsistencies are caused by four different root causes:
Incorrect programming
Incorrect error handling
Incorrect manual data entry
No clear leading system
These different root causes and the distinguishing factors are described in the next sections.

2.1.2.1 Incorrect Programming


Inconsistencies can be caused by programming errors leading to incorrect data between
different process steps or applications.
A special case of incorrect programming leading to inconsistencies is neglecting the
transactional correctness of programs. In this case, the logical unit of work (LUW) concept is
neglected, so that in the worst case only part of a business object is created or changed.
This type of root cause is very common in interfaces. A typical indicator that warrants a
detailed analysis is an inbound interface where several business steps are triggered by one
incoming document (see IF1 in Figure 2.1). If such an interface is found during the
assessment, an SQL-Trace and a code inspection should be performed to check whether
commit work statements are triggered between the different steps. If this is the case, and it
cannot be changed, the interface should update appropriate status information on which
steps have been finished successfully and should be restartable based on this status
information. Without these measures, it could happen that one step is finished successfully
but this information has not been updated in corresponding documents so that data is
created twice or not at all. In the given example in Figure 2.1, the delivery could be updated
and goods issue is posted but the sales order is not updated about this fact if a severe error
happens between step “Post Goods Issue” and step “Update Sales Order”.
SAP Active Global Support provides two services that can assist you with the detailed
investigation and improvement proposals for such situations: SAP Interface Management
Service for the technical investigation of interfaces or a Customer Program Optimization if a
similar situation is encountered outside an interface (for example, by a custom-made report
updating several documents).

6
© 2008 SAP AG
7
Best Practice: Data Consistency Check

ECC WMS
Create Sales Order

Create Delivery

Receive Delivery Information

Perform Picking

Print Labels

Perform Packing
IF1
Update Delivery

Post Goods Issue

Update Customer Specific


Table

Create Invoice

Print Billing Documents

Figure 2.1: Inbound Interface Triggering Several Business Steps

2.1.2.2 Incorrect Error Handling


Sometimes, the program code itself is correct but the operation of the system has led to
errors. This situation is most common in a CRM system or for business processes spanning
two systems. Typical scenarios for this case are:
Missing data in one of the two systems caused by BDocs that are in error state and
are not reprocessed in a timely fashion.
Overwriting newer data by incorrect reprocessing of an old IDoc.
Finding a large number of old erroneous interface data, for example IDocs or BDocs, in the
system is usually a good indicator for such a case.

7
© 2008 SAP AG
8
Best Practice: Data Consistency Check

Time t1 Time t2 Time t3 Time t4

Object A: Value1 A Object A: Value1 B Object A: Value1 B Object A: Value1 B


Value2 A Value2 A Value2 C Value2 C
System A

Replicate with
Replicate Error Replicate

Correct Error
Object A: Value1 A Object A: Value1 A Object A: Value1 B Object A: Value1 B
Value2 A Value2 A Value2 C Value2 A
System B

Figure 2.2: Inconsistency Caused by Interface Error Handling

2.1.2.3 Incorrect Manual Data Entry


Very often, the program code is correct but the system has not been used in the intended
way, leading either to a manual entry of incorrect data or to not entering important data at all.
This is most common when a new system/business process release has been introduced
and the end users have been used to the old system for a long time; this results in a
discrepancy between the physical world and system. A typical scenario would be the
introduction of a new warehouse management system release where the end users post the
stock still using old transactions or do not use some of the new transactions. The best way to
identify this root cause is to follow the daily business of one of the end users to check
whether the system usage corresponds to the designed and intended business process. In
parallel, a member of the basis team should continue to investigate the interfaces and error
handling especially if two systems are involved. The appropriate counter measure whenever
this situation is encountered is to perform another end user training with a special focus on
handling exceptions. Typical exception situations to be covered by this end user training
should be topics like “How do I proceed if I want to post material to a storage location and the
system does not allow this?” and “I receive material in an inbound delivery, which I should
not receive according to the delivery advice/has not been maintained in the system – How do
I proceed?”

2.1.2.4 No Clear Leading System


In this case, data cannot be changed in one system only, but in two systems independently.
This situation creates the risk that outdated data being sent from one leading system
overwrites new data already sent from another system. This interchangeability of data in two
systems would not pose an issue when only distinct data sets are maintained in both
systems (for example, another material type in both systems or sales data in one system and
production data in another). If you encounter a situation where the same data can be
changed in two systems, you should verify how it is ensured that no updated data (for

8
© 2008 SAP AG
9
Best Practice: Data Consistency Check
example, from MD1) is overwritten by outdated data from the other system during an
assessment. This situation is especially common for master data distribution

MD1 PD1

Upload Master Data

Create / Change
Material Data Create Extended Data

Exchange Extended
Data

MD2 PD2

Upload Master Data

Create / Change
Create Extended Data
Material Data

Exchange Extended
Data

Figure 2.3: Possible Inconsistencies by Unclear System Roles

2.2 Overall Procedure


The goal of any procedure regarding inconsistencies should be two-fold: The procedure
should identify the root cause of the inconsistency and identify and correct the inconsistent
data including dependent data. In addition, it needs to be decided quickly whether productive
use of the system should be continued, disrupted or available workarounds should be
employed. Bearing these goals in mind, the procedure outlined below has proven beneficial
in the past in providing a starting point and a guideline for a detailed analysis independent of
the individual systems and processes involved. Specific tools and detailed data models for
common occurrences are described and outlined in section 3. In addition to this procedure,
SAP has provided a Best Practice called “Business Continuity Management for SAP System
Landscapes” to minimize system impact by setting up and using a business continuity
concept.
It is common that an inconsistency is initially reported on a very basic level which does not
contain very detailed information regarding the technical data or core business processes
involved. A typical example could be a complaint by an end user that orders cannot be
delivered as the stock inventory is incorrect. Furthermore, it is usually not known how this
inconsistency has been detected. Thus, at this stage it is frequently not determined whether
there are any true inconsistencies in the system or whether only temporary differences have
been observed. Temporary differences could, for example, be caused by pending IDocs or
update tasks and are resolved once all pending system activities are completed while
inconsistencies will remain after the activities are finished. Sometimes it is even not known
whether the root cause for the described business issue is an inconsistency or difference at
all.
Taking into account the missing information at this stage, the first step should be to obtain
the missing information to reach a detailed understanding of the involved business
processes, monitoring and reporting activities, and technical objects leading to the observed
difference. The best procedure to obtain this data would be the assessment of the business
processes and the involved data flow. It may not be necessary to understand all core
business processes, but only the process involved in the inconsistency detection (which may
9
© 2008 SAP AG
10
Best Practice: Data Consistency Check
not be a core business process) as well as the core business processes which could lead to
such a reported inconsistency or will be affected by it. For example, if an inconsistency in
stock level information was reported, the process of identifying that data is inconsistent
(stock reports, error messages, and so on) as well as the business processes affecting the
stock (for example, Sales Order Management (goods issue), Production (goods issue, goods
receipt), and so on) have to be understood. While at this stage, not all core business
processes of the solution are important, it is essential to know between which business data
and technical data the inconsistency has been observed, especially for the inconsistency
detection and the business process steps handling the inconsistent data. In addition, a very
detailed, technical view needs to be obtained of the data origin and the use of the data by the
processes which are in close vicinity to the reported inconsistency, as well as any error
detection and handling around these steps.
Example: If a stock inconsistency has been reported, it is essential to know how the
inconsistency was detected/measured, how business transactions involving stock (for
example, goods movements) are executed, what data is used on a technical level (for
example, tables and fields, if several systems are involved: how data is mapped). The goal is
to understand the steps leading to the detection of the inconsistencies and to map
corresponding steps of data derivation in case different business processes are involved (for
example, stock calculation in an external reporting system and in ECC or reporting
inconsistencies between accounting in ECC and reporting in BI). An example for such a
mapping of a data derivation process can be found in Figure 2.4. The underlying data for an
inconsistency being reported between data, that was aggregated using several steps in both
systems, could, for example, be an LIS-Table in ECC (S*) for Process 1 corresponding to an
InfoCube in BW for Process 2. In such circumstances, it is important to understand which
information corresponds between the two systems.

Process1 Process2

Reported
Process Step n Process Step m
Difference

Logical
Process Step 4a Mapping Process Step 4b

Process Step 3c Log


Ma ical
ppi
ng
Process Step 3a Process Step 3b

al
Process Step 2c
gic
Lo ping
p
Process Step 2a Ma Process Step 2b

Process Step 1a Process Step 1b

Technical
Lowest (Underlying) Data Level Mapping Lowest (Underlying) Data Level

Figure 2.4: Logical Mapping of Steps Between Two Processes


Details regarding the assessment of the processes, including questions to obtain a
first working hypotheses and typical pain points can be found in section 2.3:
Assessment Guide.
10
© 2008 SAP AG
11
Best Practice: Data Consistency Check

Best Practice would be to document all this data already during the implementation
project of the core business processes and to think about possible impacts.

Once the first step of the inconsistency investigation - the understanding of the involved
business processes and data derivation processes - has been finished, the root cause of the
inconsistency and whether a real inconsistency exists have to be identified. The next step is
therefore to check for the common root causes in the steps involved directly with the
inconsistent data. The general rule here is not to dig straight into a technical analysis but first
to rule out the more operational root causes unless the inconsistency is reproducible.
A common root cause for inconsistencies is application and interface monitoring and error
handling (see chapter 2.1.2). This means that investigation of interface monitoring and error
handling should be an important part of the assessment, especially if different systems are
involved. Once the relevant interfaces and business process steps have been identified by
the recording of the involved processes, we need to determine how these are monitored
(application monitoring and interface monitoring) and whether erroneous data is contained in
the system. Besides additional, detailed questions described in the next section in table 5,
the system should be checked for old erroneous data using available standard transactions
(for example, within an SAP system ST22 for short dumps, WE02 for IDocs in error state,
and so on). The exact transactions available to check for this data depend on the interface
technique, the business process steps, and the system itself. If data in error state is found,
the handling of such incorrect data should be discussed within your monitoring team. If
incorrect handling has been identified (for example, reprocessing of BDocs or Updates
without checking that this data is still valid), appropriate error handling procedures need to be
derived.
If the reported differences cannot be explained by error handling procedures and are not
temporary differences, the logical mapping of corresponding steps including an
understanding of the underlying technical data as reached during the assessment should
now be used to verify the data consistency on the lowest technical level available (in our
example: comparison of S-table versus InfoCube). The investigation needs to be started on
the lowest level as higher level, derived data will always show inconsistencies if the lower
level data is already corrupted, even if the involved intermediate processing steps are
correct. Check reports are needed to compare this data. It is important that temporary
differences are filtered out during the use of check reports by using a time window where no
system activities are executed affecting the involved data. Filtering out differences means
that all update tasks or interface processing for the relevant data should be finished and that
no new data is created during the time of the analysis. If such a time window is not available,
the comparison has to be repeated and results need to be compared between the different
runs. Temporary differences will disappear between different runs of the check reports/tools
while real inconsistencies will remain in the result for all runs of the same report. An example
would be to compare the information between the S-Table and the InfoCube at a time where
both systems are in sync and no additional data is written to the S-table during comparison. If
this is not taken into account, the S-table may contain newer data missing in the InfoCube
which would be corrected with the next extraction.
Sometimes, the situation arises that inconsistencies exist in very old data. The old data
should be corrected before attempting to identify the root cause in these cases by reloading
or correction reports whenever possible as old inconsistencies whose root cause has
probably been fixed already may obscure the root cause for new inconsistencies or whether
inconsistencies exist at all.
Possible reports and methods to correct the outdated data can be found in section 3:
Tools and Procedures within an SAP Environment.

The business process steps leading to the low level data need to be traced and debugged
when temporary differences have been filtered out and inconsistencies have already been
identified in the low level data, as a program error is most likely, for example, the Extractor
11
© 2008 SAP AG
12
Best Practice: Data Consistency Check
filling the InfoCube and the update routines for the S-table need to be investigated in our
given example of S-table versus InfoCube. Programming errors leading to a mismatch
between both data sets could be purely technical (for example, an incorrect sign in some
formula), or logically, if the data calculation rules are incorrect (for example, using the
document date instead of a posting date). If temporary differences are ruled out and no
inconsistencies are found in the underlying data, the dependent derived data needs to be
investigated from bottom level to top level. It is important that only data mapped logically is
compared at this stage and no intermediate steps are taken into account. Using the example
for a logical mapping displayed in Figure 2.4 once the processes involved in the technical
mapping of the underlying data has been ruled out as root cause, steps 2a-2c, 3c-3b, and
4a-4b should be compared. If step 2a would be compared to step 2b, just following the order
of steps without taking into account whether a logical connection exists between the two
different data sets, the result of the consistency verification between process1 and process2
would be meaningless at this stage as data would be compared that is not comparable by
definition.
The process chain should be investigated from bottom to top using appropriate means like
check reports or queries until the first real occurrence of inconsistencies has been identified.
Once an inconsistency between two logical connected steps has been found, the technical
process steps involved between these steps need to be debugged and investigated to
identify the technical root cause like, transactional incorrect interfaces, technical
programming error, or logical error.
The root cause for the inconsistency needs to be corrected once it has been identified. The
appropriate measure to do so depends on the nature of the root cause. If a coding error was
identified as the root cause, the coding needs to be corrected. On the other hand, if the issue
was caused by the operation of the solution, appropriate measures, for example, would be to
train your Operations team (for example, using a System Administration workshop) and to
adopt your operation procedures.

Details regarding typical root causes, the distinguishing factors between these, and
appropriate solutions can be found in section 2.1.2: Typical Root Causes for
Inconsistencies.

How dependent data is affected needs to be investigated after the root cause has been
identified and corrected. This could be either consolidated reporting data or follow-up errors
like incorrectly created documents. The business processes using the original data identified
as inconsistent have to be followed to understand the impact of an inconsistency in one
business object on depending business objects. If follow-up documents are created (for
example, creation of controlling or financial data from logistics data), these need to be
corrected as well using appropriate measures.
To recover or correct lost and incorrect data sets, several possibilities exist:
Restore into a parallel system and reload of data from the parallel system by RFC,
LSMW and so on
Reload of data from a leading system
Use correction tools to recover data by relationships to redundant stored data
Manual correction
Complete restore of a backup / point-in-time recovery

The last two methods should only be used as a last resort.

Very often, a combination of data recovery methods and tools is required, for
example, individual incorrect sales documents could be corrected manually on the
data base and dependent data could be corrected afterwards by correction reports.
Each of the different methods has certain advantages and disadvantages leading to

12
© 2008 SAP AG
13
Best Practice: Data Consistency Check
different use cases.

You can find a more detailed overview about these methods in Appendix 4.3:
Advantages and Disadvantages of Different Recovery Methods.
Important questions that influence the choice of recovery method and therefore need
to be discussed are:
Does dependent data exist for the inconsistent/lost data?
Could these be used for a reconstruction of data?
How many Business Objects and how many instances are affected?
What quantity/complexity of objects is affected?
Is a backup available for a point-in-time recovery?
How much time would the different methods require?

2.3 Assessment Guide


An inconsistency reported does not necessarily have to be based on incorrect system
behavior or need to be an inconsistency at all as we have seen in the proceeding section.
The assessment given in this chapter should provide a general guideline to obtain a first
working hypothesis on what the root cause could be and a very detailed, clear understanding
of the involved business process steps and technical data which form the basis for any later
steps during root cause analysis and possible correction.
2.3.1 Understanding the Situation and Inconsistency
The starting point is always that some inconsistency is reported but the root cause as well as
the exact impact of the inconsistency and how it has been observed may not be known. To
guide the decision whether a business usage of the affected system in the current situation is
possible or not, the possible impact of a continuation versus a stop of system usage needs to
be gauged carefully based on the business impact in both cases.
The first step of the assessment should be to obtain a general understanding of the business
impact and which processes and data are affected (if this has not been done in advance as
described in the Best Practice document “Business Continuity Management for SAP System
Landscapes” ) as well as whether an inconsistency really exists.

Questions Answers
Which business objects are affected?
Describe the differences and your
expectations.
Background: Sometimes a difference
may be reported due to misinterpretation
of the data’s meaning.
Which business processes are affected?
Which business process steps are
affected?
Describe the business impact.
Background: The business impact is an
important factor to decide whether you
should continue working in the system
until the differences are resolved.
Table 2.1: Questions to Understand the Business Impact and Severity
The results of these questions determine the business processes and the business process
steps which have to be investigated in more detail. Furthermore, the determined impact is
one of the most important basis factors how business operation should proceed.
Once the severity of the inconsistency has been established and it is known which business
process are affected, the next step should be to investigate the inconsistency detection
13
© 2008 SAP AG
14
Best Practice: Data Consistency Check
process and the inconsistency in more detail. Questions to help with this task are collected in
the following table.

Question Answer
How did you notice and identify the
inconsistent behavior?
If observed by consistency check tools:
Which transactions/reports did you use?
Is the difference reproducible or did it
occur only once?
If it is reproducible:
Are the same inconsistencies reported?
How did you see that they are the same?
Did you observe any rules between
inconsistencies?
Background: If rules are observed, they
provide important information about the
nature and possible root cause of the
inconsistency. If it is observed at certain
times only, it could indicate differences
due to performance bottlenecks; if it is
only observed for certain subsets of one
business objects (for example, only
certain storage locations), this could
indicate a mapping error, and so on.
Do the inconsistencies disappear and
appear again after some time?
Table 2.2: Questions for a First Understanding of the Inconsistency
The results of these questions establish whether we have an inconsistency or only a
temporary difference between two data sets. Any rules/connections identified between the
observed inconsistencies need to be considered in the detailed business process root cause
investigations. The next questions will collect more details on the process and provide the
technical data background to verify the correctness of the data mapping and inconsistency
checking process. At this stage, data is collected to function as starting point for the technical
business process and inconsistency investigation as well as for a later correction.

Question Answer
What are the input parameters and what
did you expect?
If observed by a custom made
report/transaction:
How does the report/transaction work
technically?
Which tables and table fields are
compared?
If between different systems:
How is the data mapped?

Background: This information is used to


verify that the data is compared correctly.
Sometimes mapping of data is done
incorrectly, for example, by mapping
quantity in bin to quantity in storage

14
© 2008 SAP AG
15
Best Practice: Data Consistency Check
location or mapping document date to
posting date. You should evaluate the
meaning of table names and fields as
well as a description of the data he wants
to use for the SAP systems.
Table 2.3: Detailed Questions to Improve the Understanding of the Inconsistency

The description of data mapping needs to be used for new check tools or may be entered
directly into the generic report given in section 3.4.3 if two linked tables in one system are
involved.
2.3.2 Process Understanding
Sometimes an error in the program/interface logic or a mismatch between process design
and system usage leads to inconsistencies, especially when inconsistencies between
different systems or between the system and the physical world are observed. Thus, the
understanding of the design of the core business processes which could lead to the
observed inconsistency forms an essential part of finding the root cause, as you have to map
the process design to the system’s use in the real physical world. A detailed description of
the end-to-end process across all systems and applications has to be recorded during this
business process analysis if not documented in your company already. It is important to get
an overview of the business in general and why the chosen process has distinct priority.
For each process and process step the following information must be known:
Has a process owner been defined for this process?
Are there any other known critical problems/issues?
Short description of the process step
Important technical objects used in the process step (transactions, programs, tables)
Online/background execution of transactions

In addition to these general questions, the following questions which are especially important
to inconsistencies have to be answered. They are used to cross check whether an
operational issue (error handling, monitoring) or process design issue could cause the
detected inconsistency.

Question Answer
Has a new system or process been
introduced recently?
Did the end user training contain
instructions how to handle exceptions?
If yes: What exceptions in this area have
been included?
Have the users been used to another
system/transaction/report for similar
tasks in the past?
What testing activities (user acceptance /
integration / volume) have been
performed?
Table 2.4: Questions to Assess the System Usage
The next block of questions should be used if more than one system is involved. They are
intended to determine whether the inconsistency could be caused by several leading
systems, monitoring and error handling in the area of interface monitoring, or whether an
investigation toward transactional correctness should be started.

15
© 2008 SAP AG
16
Best Practice: Data Consistency Check
Question Answer
Is more than one system involved in the
business process steps?
If yes:
What are the roles of the involved
systems? Which one is the leading
system?
Background: Sometimes no clear leading
system is defined. In these cases, a high
effort is needed to keep data consistent
between the systems as changes in one
system could be overwritten by changes
in another system. Common cases are
master data development systems.
Describe the relevant business process
steps affected by the observed
inconsistency
Are interfaces to other systems involved
in these steps or in prior steps regarding
the used business data?
If yes, how do you perform reconciliation
between the involved systems?
Background: Sometimes the temporal
order of reconciliation reports does not fit
the timing of the business process steps.
For example, when running an MRP in
an external legacy planning system, the
current requirements and stock need to
be synchronized with the SAP ERP
backend.
What interface technology do you use
between the two systems?
Is custom-made coding used?
Do you trigger more than one step by
one call of the interface?
What measures have been taken to
ensure that the interfaces are
transactionally correct?
Please describe your monitoring and
error handling concept?
Background: Especially if more than one
step is triggered by an interface using
custom code, the transactional
correctness could be compromised.
Even when the transactional correctness
is ensured, incomplete monitoring and
error handling could lead to data loss.
Using the example of running MRP in an
external legacy planning system,
unobserved errors in the transfer of
requirements and stock will lead to
incorrect planning results due to
inconsistencies between the systems.
Hint: If one of the first two questions is
16
© 2008 SAP AG
17
Best Practice: Data Consistency Check
answered yes, you should involve an
interface or ABAP expert skilled with
transactional correctness investigations
in the next steps or order an Interface
Management Service.
Table 2.5: Questions for Processes Involving Several Systems

17
© 2008 SAP AG
18
Best Practice: Data Consistency Check

3. Tools and Procedures within an SAP


Environment
Tools to find and correct inconsistencies within an SAP environment will be provided within
this section. These include standard tools as well as commonly used tools that are provided
in SAP notes only and must be implemented in customer namespace. Only the most
important ones will be described in this section.

3.1 Tools to Check Consistency Between ECC/CRM and Within


CRM
An SAP CRM System usually consists of more than one logical database. Every SAP CRM
System has a CRM database. In most cases, data exchange with one or more R/3 back-end
systems is necessary. The consolidated database (CDB) is the basis for data exchange with
Mobile Clients. It is very important to keep these databases synchronized and consistent,
which is the task of the SAP CRM Middleware. However, inconsistent data may occur if:
The Delta Load was inactive for some time. For more information about Delta Load
refer to Activating Object Classes for Delta Synchronization in the CRM online
documentation.
The Initial Load could not be carried out completely. For more information about the
Initial Load, refer to Starting the Initial Data Transfer in the CRM online
documentation.
Data was accepted in the CRM System but rejected in an R/3 back-end.
Filter conditions were changed for the load from an R/3 back-end. For more
information about filter conditions, refer to Defining Filters for Objects in the CRM
online documentation.
To analyze and repair inconsistencies between SAP CRM and SAP R/3, SAP created the
tool DIMa (Data Integrity Manager). The tool is used to analyze and correct inconsistencies
resulting from the exchange of the most important business objects between R/3 and SAP
CRM. The features of DIMa and covered business objects are described in the next section.
Furthermore, SAP has created various analysis and correction reports for the SAP CRM and
R/3 exchange scenario and also for the SAP CRM Mobile scenario. The most important of
these reports are described later in this document.
3.1.1 Data Exchange for the Most Common CRM Scenario - CRM Sales
The most commonly used CRM scenario is CRM Sales. In this scenario, sales orders and
other objects are exchanged between SAP CRM and R/3. Inconsistencies of business
objects may occur due to errors in the exchange process. The DIMa is a good tool to
uncover inconsistencies between business objects in R/3 and SAP CRM. The data exchange
toolbox (note 718322) provides various tools which are described later to analyze these
inconsistencies in more detail.
Be aware that inconsistencies can be temporary. Business objects are exchanged
asynchronously in queued RFCs of the middleware. If you detect a difference for a
business object, the changes that rectify the difference might already be in the
queues but not yet processed by R/3 or CRM. On the next run of DIMa, this temporary
difference would not occur anymore as the changes would be processed from the queues.
Persisting differences after several runs of DIMa can be considered as real inconsistencies.
Usually, if you detect errors (inconsistencies in DIMa) in the exchange, for instance, of a
sales document created in SAP CRM and distributed to R/3, you often run into the error that
R/3 rejects the created sales document. The document exists in CRM but does not exist in
R/3. A correction, for instance, customizing in R/3 has to be performed to enable successful
reprocessing of the rejected sales order. (The order can be resent by an action in the
CRMD_ORDER transaction in SAP CRM).

18
© 2008 SAP AG
19
Best Practice: Data Consistency Check
Also, errors in middleware like items deleted in queues can result in inconsistencies between
CRM and R/3. To deal with such inconsistencies proactively, interface monitoring is
advisable.
As CRM often uses filters between R/3 and CRM, special care must be taken, as changing
these filter conditions can result in inconsistencies. For instance, if the condition on R/3 side
is widened, business objects may be sent as delta messages which do not have a
corresponding object on CRM side. If, for instance, filter conditions are widened on the R/3
side, an initial load of the respective business objects must be performed.
3.1.2 Data Exchange for the CRM Scenario CRM Mobile
In the CRM scenario CRM Mobile, laptop users (mobile clients) exchange data (business
objects) with a central CRM system via a complex middleware. In this scenario, the CRM
system has a special database for mobile client data processing: the CDB (consolidated
database). Inconsistencies can occur between the CDB and the central database of the CRM
system as well as between the CDB and the mobile clients. The DIMa can be used to
analyze inconsistencies between the central database of the CRM system and the CDB.
Be aware that inconsistencies in a CRM environment can be temporary. Business
objects are exchanged asynchronously in queued RFCs of the middleware. If you
detect a difference for a business object, the changes that rectify the difference might
already be in the queues but not yet processed by CRM. On the next run of DIMa, this
temporary difference would not occur anymore as the changes would be processed from the
queues. Only persisting differences after several runs of DIMa can be considered as real
inconsistencies.
Two common cases of inconsistencies exist between the mobile clients and the CRM server.
1. Data is missing on mobile clients but is present in the CDB. In this case, the
administrator corrects the inconsistency by performing an extract of the business
object in question using transaction SMOEAC for the respective mobile client.
2. A more serious inconsistency is that data changed on the client is not processed
properly by the CRM server because the server rejects the change. These
inconsistencies are only detected in rare cases as the respective BDocs have the
status finished (F02). To detect such inconsistencies, the system administrators
should monitor for BDocs with state F02 in SMW01/SMW02 transaction. As well on
the mobile clients, rejected objects can be found by navigating to the CRM inbox
tileset. If a rejected change is detected, it can be further analyzed why the rejection
occurred. Then, either correction at CRM server level or corrections on the mobile
client have to be performed in order to reprocess the BDoc successfully.
As the outbound queues are sometimes huge, especially after rollout of mobile clients or
mass updates triggered in R/3, outbound queues are sometimes deleted which results in
inconsistencies. To correct these inconsistencies, an extract of the business objects for the
respective clients has to be performed.
3.1.3 DIMa – The Tool to Detect and Repair Inconsistencies Between
SAP ECC and SAP CRM
The Data Integrity Manager (DIMa) helps to detect and repair inconsistencies between
databases in the SAP CRM system, the R/3 backend and SAP CRM’s mobile part. It
supports consistency checks for the most commonly used business objects: business
partners, products (materials), sales documents and pricing information. The DIMa uses the
SAP CRM middleware and the CRM R/3 plug-in to carry out the extraction and comparison
of business objects. DIMa also provides the possibility to correct inconsistencies in addition
to detecting inconsistencies.

3.1.3.1 Features of DIMa


The DIMa compares data in different databases and displays inconsistencies. The data
comparisons can be carried out between:
The CRM database and an R/3 back-end database
19
© 2008 SAP AG
20
Best Practice: Data Consistency Check
The CRM database and the CDB database.
For many objects, it is not only possible to compare but also to synchronize the data via
DIMa. There are two compare types available in the DIMa:
Header Compare:
A header compare checks, whether an object instance exists in both databases.
Detail Compare:
A detail compare compares all data of an object instance found in both databases.
Some objects may not allow a header compare. The detail compare is then carried out.

3.1.3.2 Consistency Check Scope of DIMa


DIMa supports consistency checks for the following business objects:

DIMa check object Business object Starting from CRM Rel., SP


ACT_DIMA DIMa Object for Activity 4.0*
BP_DIMA DIMa Object for BP 4.0*
CAM_DIMA DIMa Object for Campaign
CDB_CAPGEN Bus. partner CRM CDB 5.0
CDB_CONGEN Contact pers. CRM CDB 5.0
CONTACTS R/3 Customer Master: Contacts 3.0
CP_DIMA DIMa Object for Contact Person 4.0*
CUSTOMER R/3 Customer Master: Customer 3.0
CONDITION TABLES Condition tables R/3 CRM 4.0
EMP_DIMA Compare Employee 4.0*
MATERIAL Material: ERP => CRM 4.0; detailed 5.0
OPP_DIMA DIMa Object for Opportunity
PARTNER R/3 BP Master: Partner 3.0 SP14; 3.1 SP04; 4.0;
detailed 5.0
PRODUCT_MAT Material: CRM => CDB 3.0; detailed 4.0
PRODUCT_SRV Service: CRM => CDB 3.0; detailed 4.0
REBATE_DNL_AGR Agreements R/3 CRM 5.0
RELATIONS R/3 BP Master: Relations 3.0 SP14; 3.1 SP04; 4.0;
detailed 5.0
SAD_DIMA Salesdocument CRM -> CDB
SALESDOCUMENT Documents R/3 --> CRM 3.0 SP14; 3.1 SP04; 4.0
SAMPLE_CDB Sample for CDB CRM
SAMPLE_R3 Sample for R/3 CRM
SERVICE_MASTER Service: ERP => CRM 3.0; detailed 4.0
Table 3.1: DIMa Check Objects

For further details about the use of DIMa, refer to the DIMa documentation and the online
help.

You should execute DIMa for the most important business objects on a regular
basis. The monitoring object “DCCRMDIM” may be used to monitor these regular
runs within the SAP Solution Manager.

3.1.4 Analysis and Correction Reports in Data Exchange of One Order


Documents Between SAP CRM and R/3
SAP created a data exchange toolbox report to investigate the exchange of one order
documents in the CRM middleware. This toolbox is provided by composite SAP note 718322
which includes the most important analysis and correction reports regarding data exchange
of one order objects between CRM, R/3 and CRM Mobile.
20
© 2008 SAP AG
21
Best Practice: Data Consistency Check
An often used analysis report is CRM_ODE_CHECK_CFG (SAP note 718797). It checks the
configuration of the data exchange between SAP CRM and SAP ERP/R/3, both on the CRM
and the R/3 side. Information like the RFC destinations for the R/3 system are displayed as
well as the queue prefixes of the queues relevant for the data exchange between CRM and
R/3 on the CRM side. The RFC connections in the opposite direction are displayed on the
R/3 side, as well as various information about the configuration, like the queues involved in
the exchange.
A very important report in root cause analysis of issues with one order objects is
CRM_ORDER_READ. This report displays on the database level all the information linked to
a CRM one order object. Often, this report is the initial starting point in analyzing the root
cause for problems with one order objects. A one order object saved in a CRM system with
errors is not replicated to R/3. Sometimes, users want to have one order objects replicated
although errors exist. The report CRM_MESSAGES_DELETE deals with this issue (SAP
note 719224). The errors are deleted by the report from the application log for the one order
objects in question to enable the replication into R/3. The affected orders are replicated by
saving the one order objects in the report.
The central SAP note for all the reports in the data exchange toolbox is 718322. A list of all
the reports and a short description follows:

SAP Note Short Description Report


718751 Read Tables Business Transaction CRM_ORDER_READ
Root cause analysis: Display data structure of a business transaction in detail
718797 Check Basis Configuration Data Exchange CRM_ODE_CHECK_CFG
Configuration of data exchange between R/3 and CRM (RFC destinations, and so on)
718916 Set Message Log Level CRM_MESSAGES
Set Log Level in CRM for your terminal session / Set breakpoints for certain messages produced
718918 Display BDocs for a Business Transaction CRM_DATAEXCHANGE_1O_BDOCS
Display m/sBDocs for an business document (sales order, activity, and so on) in CRM
719032 Compare a Document in CRM and R/3 CRM_ODE_CHECK_DOC
The report checks whether a transaction is available in R/3 and CRM if both if they are unequal.
718929 Incorrect Documents that were created in CRM_DATAEXCHANGE_CHECK_OLTP
R/3
Troubleshooting for documents that were exchanged between R/3 and CRM but resulted in an error
on CRM side
Table 3.2: Analysis Reports

SAP Note Short Description Report


648621 Checks after initial data transfer if ZZ_DOWNLOAD_REPAIR_ERRORS
incorrect documents exist in CRM
Checks after the initial data transfer, if incorrect documents exist in the CRM server, deletes them in
the CRM server and retrieves them from the R/3 system again.
686063 Check after initial data transfer, if all CRM_INIT_DNL_MERGE_ORDERS
documents have arrived in CRM
Checks after the initial data transfer, whether all documents have arrived in the CRM server and
retrieves possibly missing documents from the R/3 system.
717478 Deleting a CRM OneOrder Document CRM_ORDER_DELETE
Deletion of one order objects in CRM. Subsequent deletion in CRM Mobile, R/3 can be omitted.
719109 Creating Missing Replication Records CRM_CREATE_REPL
When creating one order objects in CRM, they are uploaded to R/3 and return as replication records
for the one order objects. Sometimes, these replication records get lost. To create these confirmation
records use this report.
21
© 2008 SAP AG
22
Best Practice: Data Consistency Check
719111 Setting Completed Status for a Sales CRM_SET_STATUS_COMPLETED
Document
Set the status of one order objects exchanged between R/3 and CRM in CRM to ‘completed’. Not
completed may be caused in CRM for various reasons.
719113 Delete Links to PRCD_HEAD CRM_LINK_OBJTYPE_18_DELETE
CRM 3.0/3.1: Wrong pricing documents due to program error for CRM order documents. Correction
report deletes wrong pricing documents. (see also 579336/599970)
719224 Delete Messages from Error Log CRM_MESSAGES_DELETE
Error messages in application log are deleted for a certain one order document.
719225 Delete Unnecessary Replication Records CRM_REPL_CHECK_AND_DELETE
Replication records exist in the CRM server for which no more CRM documents exist. This report is
used to delete these orphaned replication objects. See also note 719109.
719227 Delete Orphaned CDB Entries CRM_ORDER_DELETE
You used report CRM_ORDER_DELETE to delete documents in the CRM server and selected the 'Do
not send BDoc type' flag. As a result, the CDB entries were not deleted as well. (See 717478)
719229 Contracts from R/3: Adapt Doc. Flow CRM_CONTRACT_UPDATE_R3
Document flow records and release values for contracts which were loaded from the R/3 system to the
CRM server are incorrect. See also note 736044.
719960 Resend CRM Documents CRM_START_REQUEST_FROM_CRM
During the data exchange between the CRM server and the R/3 system, documents can only be
transferred to the R/3 system by means of a change in the CRM server. A request download from the
CRM server to the R/3 system does not exist.
721107 Compare delivery - and Bill. Status CRM CRM_R3_CHECK_STATUS
between R/3 and CRM
You can use the CRM_R3_CHECK_STATUS correction report to read the delivery and billing status
from the R/3 system and to adjust the status correspondingly in the CRM system.
Table 3.3: Correction Reports

3.1.5 Miscellaneous Check, Correction and Repair Reports in CRM


3.1.5.1 Inconsistency Check and Correction Reports
An important correction report is CRM_ORGMAN_DOWNLOAD. Every time you change the
organization model in CRM online, you have to run this report to update the organizational
changes in CRM Mobile (CDB). Please refer to the documentation in the customizing
transaction SPRO CRM -> Master Data -> Organizational Management -> Data transfer ->
Copy Organizational Structure to CRM Mobile.
When applying SAP notes for index access optimization, often new fields are introduced in
the CRM_ORDER_INDEX table for instance. After introducing such fields, the
CRM_ORDER_INDEX is inconsistent as the new fields do not hold the necessary data. To
repair the inconsistency, run CRM_INDEX_REBUILD to rebuild the whole table or for
instance report CRM_INDEX_FAST_TEMPL_TYPE_UPD to populate only the necessary
new fields.
For the CRM Mobile scenarios, certain temporary inconsistencies between mobile client and
the CRM server can be avoided by setting specific entries in the table SMOHNOCSTA on the
CRM server. The temporary inconsistencies can occur if you perform the following:
First, you make changes to a business object on the mobile client and you upload them to
the CRM server by ConnTrans. Secondly, you proceed after the ConnTrans to change the
business object again. After performing a second ConnTrans, the changes you just made are
reset by the server to the state after the first ConnTrans. These changes are updated
correctly on the client to the state of your last change after a next or later ConnTrans run.
To avoid these temporary inconsistencies, changes in the SMOHNOCSTA table on the CRM
server are necessary, for further details refer to SAP note 845693.

22
© 2008 SAP AG
23
Best Practice: Data Consistency Check
As well for the CRM Mobile scenario, internal inconsistencies in look up tables or extract
queues for intelligent and interlinkage objects can be corrected by using the report described
in SAP note 622693. The corresponding correction report is
Z_30/40_EXTRACT_LUTAB_EXTRACTED_F. Refer to the SAP note 622693 for further
details.

3.1.5.2 Root Cause Analysis Reports


A report for analyzing the Mobile Client distribution model is ZTN_RR_LU_ANALYSIS. This
report is very useful for analyzing the distribution model of the mobile clients to uncover
incorrect configuration in the distribution model. The most important functionality is the
possibility to analyze the distribution model of intelligent objects. Refer to SAP note 835149
for further documentation.

3.2 Tools to Check Consistency Between ECC/SCM and Within


SCM
3.2.1 Introduction
The term Internal consistency refers to the consistency between the liveCache data and the
APO database while the term External consistency refers to the consistency between the
SCM system and one or more dedicated R/3 Systems. Please make sure that the internal
consistency of the APO system is guaranteed before you check the external consistency and
set it up again.
All check reports provided by official SAP notes in the area of SCM can be executed by the
system administrator or scheduled or manually by the end-user in update mode.
3.2.2 Product Allocation (PAL): Internal Inconsistencies Within SCM or
Between ECC/SCM
The functionality of SAP SCM PAL is very flexible to use, however, expectations to its
functionality are usually very specific and sometimes differ from the standard design. Several
user-exits are available to adapt the functionality to specific needs.
SAP did a fine-tuning of the accuracy of product allocation in 2004 and several notes were
created. At the least, all those notes from 2004 should be applied when using product
allocation. Specific newer notes should be applied to each R/3 and SCM system when using
product allocation in addition to these notes.
Generally, the probability that frequent PAL inconsistencies in a system are due to a
customer modification or user-exit is very high when the above mentioned notes are applied.

3.2.2.1 Correction Report /SAPAPO/RMQUOT_USAGE_CHECK


Correction report /SAPAPO/RMQUOT_USAGE_CHECK (corresponding transaction is
/SAPAPO/ATPQ_CHKUSG) checks the consistency and corrects product allocation (PAL)
assignment within SCM. You can find this report in the SAP Menu: 'Advanced Planning and
Optimization -> Global ATP -> Environment -> Product Allocations -> Repairs'. See SAP note
676128 for further information on this report.

3.2.2.1.1 Check Frequency and Scheduling


Run this report at least weekly during low system activities affecting product allocation, or if
needed, once a day. Note that this report has to be started online and requires manual
interaction to correct inconsistencies. The report is often scheduled in the background, to
evaluate the spool lists only. In case of errors, it is started manually.
It is possible to run this program in the background with an automated update of the checked
items by the following workaround: Run the Batch Input for the program
/SAPAPO/RMQUOT_USAGE_CHECK using SAP Menu: System -> Services -> Batch Input
-> Recorder. Perform all the steps and then schedule it.
23
© 2008 SAP AG
24
Best Practice: Data Consistency Check
3.2.2.2 Correction Report /SAPAPO/SDRQCR21 with Product Allocation
/SAPAPO/SDRQCR21 can compare and correct the product allocation assignments across
R/3 and SCM (tables: QTVB in R/3 and table /SAPAPO/SDQTVB in SCM), if the flags and
fields for product allocation are used.
Table QTVB can be interpreted as a backup in case PAL data is lost in SCM. Therefore, this
comparison is usually used as a data recovery from ERP, if data in SCM was lost.

Note that this report does a neutral PAL check and an over confirmation might occur
that should be taken care of by a BOP run as a second step. For further information,
see the F1 help in the report /SAPAPO/SDRQCR21 for the PAL checks.

3.2.2.3 Typical Examples of PAL Inconsistencies


There are some typical occurrences of product allocation inconsistencies:
Incoming order quantity is greater/smaller than the product allocation assignments in
a period
Negative values with product allocations (which makes no sense at all):
Negative order incoming quantity
Negative product allocation assignments
Example:
The Result list of the product allocation check report in SCM found in this case a negative
product allocation assignment:

Figure 3.1: Result List for Product Allocations


Such negative PAL confirmations can also be found in table /SAPAPO/SDQTVB:

Figure 3.2: Corresponding Table View


PAL errors with “negative confirmations” do not correspond to any economical concept and
are meaningless. Possible root causes for such situations are:
Missing SAP notes fixing a bug
User-exits
Modifications

24
© 2008 SAP AG
25
Best Practice: Data Consistency Check
3.2.2.4 Overview of Notes for PAL Consistency:
Component Sap Note Description
SCM-APO- 676128 Product allocations: Control of Product allocation
ATP-BF-PAL assignment
SCM-APO- * All *PAL notes from year 2004: note search with ERP
ATP-BF-PAL AND SCM patchlevel for “product allocation”
Table 3.4: Notes for PAL Consistency

3.2.3 Time Series of DP and SNP


3.2.3.1 Reports to Check Consistency within this Area
/SAPAPO/TS_LCM_CONS_CHECK performs a consistency check for time series of DP and
SNP for a selected planning area. See SAP note 425825 for further information on this
report.
/SAPAPO/TS_LCM_CONS_CHECK_ALL does the consistency check for all time series but
can only display the results, without correcting errors.

3.2.3.1.1 Check Frequency


It is recommended to check the consistency of time series once a day.
3.2.4 Shipping and Vehicle Scheduling
/SAPAPO/VS_CONS_CHECK detects incorrect customizing settings for the shipment
creation process. This includes inconsistencies between SCM and R/3 settings (integration
customizing), optimization profiles, inconsistencies within the different database tables and
SAP APO master data that is relevant for TP/VS, for example vehicle resources,
transportation lanes, and locations.
If inconsistencies are found, you can correct such errors by forward navigation. See SAP
note 425825 for further information.

3.2.4.1 Check Frequency


Run the report /SAPAPO/VS_CONS_CHECK once a day if needed.

3.2.4.2 Important Notes for TP/VS Consistency:


Component SAPNote Description
SCM-TEC 425825 Consistency checks, …
Table 3.5: Important Notes for TP/VS Consistency

3.2.5 Integration Models (External Consistency)


The following reports run for example daily in the R/3 / ECC system:
Program RAPOKZFX
This program checks and adjusts the field MARC-APOKZ to keep it in
synchronization with integration models.
Report RCIFIMAX
This report generates and reconciles runtime versions of integration models.
3.2.6 Internal Consistency Check for SCM regarding LC <-> DB
Transaction /SAPAPO/OM17 is the main tool for an overall internal consistency between
liveCache and APO-DB.

SAP recommends that you perform consistency checks during a posting-free period
of time. If you cannot do this, it is possible that inconsistencies between liveCache
and the SAP APO database that existed only briefly will be displayed.

25
© 2008 SAP AG
26
Best Practice: Data Consistency Check
You should re-check the inconsistencies that were displayed (Ctrl+F2: Button “Check again”)
in this case.
If the inconsistencies are still displayed after the check, you can assume that the
inconsistencies did not just exist temporarily and use the appropriate transaction to correct
them. It is recommended to execute the program in the background (F9 or Menu Path
Program -> Execute in Background as it is possible to use the functionality “Evaluate Last
Background Job” afterwards. See SAP note 425825 for further information on this report.
Several consistency check reports were integrated in this transaction as of APO release 3.1
like transaction /SAPAPO/REST02 for resources. During this reimplementation, the scope
has been enhanced so it is recommended to use the functionality within /SAPAPO/OM17.

3.2.6.1 Check Frequency


The different business objects should be checked once per day if needed.

The monitoring object “DCAPOM17” exists within the SAP Solution Manager’s
Business Process Monitoring functionality to facilitate the regular monitoring with
OM17

3.2.6.2 Important Notes for LC<->DB Consistency:


Component SAP Note Description
SCM-TEC 425825 Consistency checks, …
Table 3.6: Important Notes for LC <-> DB Consistency

3.2.7 Consistency at Interface: Transaction Data (External Consistency)


There are several tools to monitor the inbound and outbound queues across the systems.
Activate and use CIF post processing (transaction /n/SAPAPO/cpp1 or report
/SAPAPO/CIF_POSTPROCESS) or otherwise use the qRFC monitors/SCM Queue
Manager/Core Interface Cockpit to monitor issues with blocked queues to identify master
data problems, correct master data, and retrigger the queues.
External consistency checks like delta report3 will display inconsistencies resulting from
blocked queues and resend the data. However, this will not correct the inconsistencies as the
queues are still blocked.
By just deleting blocked queues, delta report3 would correct the hereby provoked
inconsistencies but such a procedure is not at all recommended. You should identify
the root cause of the blocked queue and retrigger the queue processing instead.

You could utilize proactive online monitoring and alerting within the SAP Solution
Manager’s Business Process Monitoring Framework besides the reactive CIF
specific monitoring. The appropriate monitoring object for qRFC monitoring is
IMQRFCMO

3.2.7.1 Example
CIF post processing displays SCM inbound queues that cannot update SCM. The queues
relate to VMI Sales Order 4185622/10, 4185622/10 and 4186738/20. The delta report
detects this error:

Figure 3.3: Example for External Inconsistencies

26
© 2008 SAP AG
27
Best Practice: Data Consistency Check
Reactivating the queues, the following error messages are found in the SCM application log
in transaction /SAPAPO/c3: “Location 0001065469 does not exist”:

Figure 3.4: Application Log


A comparison with the VMI order header leads to the finding that a new R/3 Ship-to-party
was created which is used with VMI orders.

Figure 3.5: Details of the Affected Order


The reason was therefore a master data problem and the new Ship-to-party was not yet
transferred to SCM.
3.2.8 Temporary Quantity Assignments (TQAs)
3.2.8.1 Root Causes of Old “Expired” TQAs and Business Impact
Typical root causes of old non-persistent TQAs are:
Manually deleted queues (for example, because of unresolved “sysfail” status)
Manually deleted transactional RFCs in transaction SM58
Time out dumps in R/3 (monitoring and RC analysis with transaction ST22)
Update errors in R/3 (monitoring and RC analysis with transaction SM13)
Issues with mass delivery processing (transaction VL10)
Back order processing (transaction /SAPAPO/BOP)
The business impact of incorrect temporary quantity assignments are sales orders which will
not be confirmed during manual GATP checks, interactive backorder processing (BOPI), and
backorder processing (BOP), even if stock is actually available (ATP under confirmation).
If deleting required TQAs manually, ATP over confirmations (= negative ATP) can occur.

3.2.8.2 General Monitoring and Tools


You should ensure that the temporary quantity assignments you want to delete do not
originate from processes that have not yet finished. Therefore, it is essential to check
whether there are corresponding queue entries using transaction SMQ2. If so, these must be
27
© 2008 SAP AG
28
Best Practice: Data Consistency Check
restarted. Since locking problems are often the reason why temporary quantity assignments
are kept, you should check transaction SM58 regularly and process transactional RFCs if
necessary. This transaction should be scheduled as a batch job (for example, every 15
minutes) by scheduling report RSARFCEX as a job.

3.2.8.3 Check Frequency


The deletion of old temporary quantity assignments should be performed regularly by
scheduling the report /SAPAPO/OM_DELTA_REMOVE_OLDER at least once (or, if
required, several times) per day as a general recommendation. The definition of the deletion
offset depends strongly on the customer’s business scenarios and monitoring procedures.
Examples are 1 hour, 1 day or 2 days. In addition, if required, the TQAs should be monitored
regularly using transaction /SAPAPO/AC06 and deleted manually on demand.
TQAs caused by backorder processing are deleted using transaction
/SAPAPO/BOP_DELETE. But the BOP result lists may be kept longer in the system. See
also SAP note 488725 for further information.

3.2.8.4 Overview of notes for TQA consistency:


Component SAP Note Description
SCM-APO-ATP 488725 FAQ: Temporary quantity assignments in Global ATP
Table 3.7: Notes for TQA Consistency

3.2.8.5 Tips for Root Cause Analysis of Erroneous TQAs

3.2.8.5.1 Use Central System Log


Larger installations often consist of many application servers, whereby the system log is
written for each single application server. To reduce the time needed to monitor each
application server using transaction SM21, it is possible to use central logging choosing the
menu path: System log -> Choose central system log / all remote system log.

3.2.8.5.2 Use of Transaction /SAPAPO/AC06


1. Define a layout (Ctrl + F8) that contains the additional displayed column “Generation Date”
and save it.

Figure 3.6: Defining the Screen Layout


28
© 2008 SAP AG
29
Best Practice: Data Consistency Check
2. Set a filter in transaction /SAPAPO/AC06 by marking PerIndTQA, right click and select
“Set filter…” by Persistency Indicator = N (non-persistent):

Figure 3.7: Defining a Filter


Comment on the “Generation Date”:
Do not use the fields “gen. date / time (DDIC)”, as they refer to the local time zone of the
end-user while dumps and the system log are written in system time which differs very often
from the time zone of the end-user. The generation date of TQA is given in system time =
UTC. Generally, you can compare the system time and time zone by selecting menu path:
System -> Status.

Figure 3.8: Comparison of Time Zones

The main issue that makes it very complicated to find the root cause of non-cleared TQAs is
that the originating transaction cannot be seen in monitoring transaction /SAPAPO/AC06 but
only the transaction GUID. The transaction GUID (TRGUID) is a key for temporary objects
that have been created during an availability check. All these objects have the corresponding
transaction GUID and can therefore be identified by it. The transaction GUID is not saved in
R/3 and linked to the transaction in a special table.

Figure 3.9: Investigation of TQAs


The following checks are recommended based on this data:
1. Take the end-user name (field Name) and look for update errors in transaction SM13
by selecting this user and flag “terminated” only.
2. Check the system log in transaction SM21 by central logging for this user
29
© 2008 SAP AG
30
Best Practice: Data Consistency Check
3. Check dumps in transaction ST22 for this user
4. Use the time stamp of the TQA (field Generation Date) and check whether the
identified potential issues match this point in time
5. Check the affected document (field Order) and change history in the corresponding
application transaction (for example, va03: Menu path: Environment -> Changes): If
the “suspicious” changes were executed by a background job, you can find the job by
searching in transaction sm37, using the button “Extended job selection” and filtering
for jobs that were running around the time of the document changes (for example, in
the depicted case from 07:02:15 till 07:02:20)

Figure 3.10: Verification of Change Logs


6. Check whether related BOP runs exist by searching with the Transaction GUID
(TRGUID) in table /SAPAPO/BOP or /SAPAPO/BOPHEAD using transaction se16

Figure 3.11: Verification of /SAPAPO/bophead using the Data Browser


In case of regular TQA issues with deliveries, go to transaction VL10 and check if there are
delivery runs by user name (field “Name”) that can match also with (field “Generation Date”).
If you find no related error messages in the VL10 log, you can open a customer support
message for SAP development.

3.2.8.6 Root Cause Analysis of Erroneous TQAs: Examples


To resolve the root causes of inconsistencies, you need to use monitoring to identify typical
and reoccurring errors. The main challenge is to reproduce these errors. Reproducing the
errors is the only way to correct modification errors.
These modification errors can then be resolved either by your internal development
department or by SAP development but might be charged as remote consulting.
30
© 2008 SAP AG
31
Best Practice: Data Consistency Check
3.2.8.6.1 Example
In the case of many TQAs linked to one transaction GUID with the ATP category BR =
deliveries and one user, it is likely that the root cause is the collective delivery runs
performed by the same user:
This case was monitored with /SAPAPO/AC06: Note the large transaction GUID for the
same Generation date 08.11.2006 and Generation time 15:24:54 till 15:25:00 by the same
user. (The ATP category BR = deliveries is not visible in this screenshot).

Figure 3.12: Overview of TQAs


Select the VL10 collective processing logs and select by user and date:

31
© 2008 SAP AG
32
Best Practice: Data Consistency Check

Figure 3.13: Displaying the Log in Transaction VL10


Find the logs of the delivery runs that generated the TQAs:

Figure 3.14: Display of the Log-file


32
© 2008 SAP AG
33
Best Practice: Data Consistency Check
Check the errors (you can see the number of errors displayed: for example, 40 errors) and
navigate to the errors. If those errors cannot be associated with TQA issues or there are no
errors in the log, a trap code could be discussed with SAP development and implemented in
order to find the VL10 issues when related expired TQAs remain.
3.2.9 Transaction Data: All Order Types and Stock (External
Consistency)
It is recommended to use the /SAPAPO/CIF_DELTAREPORT3 to compare/reconcile
transaction data and correct data inconsistencies between SAP APO liveCache and SAP R/3
DB. Data inconsistencies can occur during the transfer using SAP APO Core Interface (CIF),
if, for example, you delete entries in the queue manually or if faulty queue entries exist. The
CIF comparison/reconciliation takes into account both objects that are not available in one of
the systems and objects that differ in the two systems.

The regular monitoring of this report may be facilitated with the monitoring object
“DCAPOCCR” in the SAP Solution Manager’s Business Process Monitoring
Framework.

3.2.9.1 Handling of Temporary Differences

3.2.9.1.1 No Iteration Functionality Applied Yet?


You can determine whether inconsistencies exist between the SAP R/3 DB and SAP APO
liveCache by checking the job log, and therefore if the report has to be restarted in online
mode. This should only be the case for old releases! See SAP note 425825 for details.

3.2.9.1.2 Iteration Functionality (Online and Background)

It is strongly recommended to apply SAP notes 496779 and 488747, which introduce
the iteration functions for the CIF_DELTAREPORT3.

Due to long delta report runtimes, objects may have been changed and may no longer be
inconsistent. Temporary data inconsistencies occur during the transfer of data between SAP
APO and SAP R/3 and disappear again after the transfer. Note, however, that in the case of
lengthy transfers, inconsistencies may still be displayed as errors.
You should use the iteration for the online comparison and for the comparison in the
background, as well as when saving and loading results.
The user should interactively compare the displayed incorrect objects from the result list
again (iteratively) after the DELTAREPORT3 has performed a comparison and displayed the
result when using the online comparison. If you execute the iteration several times in a row,
you can further minimize the number of probable errors caused by temporary data
inconsistencies.
You should set the indicator Iteration for the comparison in the background where
intermediate results are saved and loaded for comparison. Temporary errors or already
solved errors are cleansed and do not appear in the result list anymore. The update of the
errors is triggered manually as the next step.
You can also use the iteration after the reconciliation to check whether the correction was
successful or not. In this way, you can determine whether an error still exists after the
reconciliation or not.

3.2.9.2 Recommendation
When using the /SAPAPO/CIF_DELTAREPORT3 for sales orders, setting the flag “Use
Table VBBE for Sales Order Comparison” is recommended for a much better performance.
You have to run report SDRQCR21 on the SAP R/3 / ERP side in simulation mode first (see

33
© 2008 SAP AG
34
Best Practice: Data Consistency Check
also paragraph 3.3.1.5) to ensure that the requirements situation is correct in the primary
system if this variant is selected.
If incorrect requirements exist, run SDRQCR21 again with data transfer before executing
/SAPAPO/CIF_DELTAREPORT3 on APO/SCM side. Inconsistencies from ERP are likely
transferred to SCM generating a business impact if you skip this step. The BOP runs could
be executed on a defective data basis resulting in ATP over confirmations (= negative ATP)
or ATP under confirmations.

3.2.9.3 Check Frequency


The different business objects should be checked by CIF comparison/reconciliation of
transaction data at least once a week up to once per day if needed.

3.2.9.4 Example for Inconsistencies Detected by the CIF


Comparison/Reconciliation
In this example, the result overview of the delta report is displayed. Checked were 750 sales
orders and deliveries across a SAP SCM and SAP ERP system:

Figure 3.15: Result Screen for CIF Delta Report


The delta report detected 142 Sales orders in SCM that are not in the R/3 system anymore
and 149 sales orders missing in SCM (error: Missing in APO). That kind of errors is always
critical and might have a business impact.
Typical root causes for both errors are:
VBBE entries were deleted or generated due to temporary (fake) errors with
SDRQCR21
Queues were deleted
The report detected additional 493 errors of type “Differences in Content”. This kind of errors
might not be critical depending on the difference in content.
Typical root causes for this kind of error are:
Queues were deleted
User-exits regarding the detailed scheduling of an order item
Bugs

34
© 2008 SAP AG
35
Best Practice: Data Consistency Check

3.2.10 Sales order requirements and SD mapping tables (external


consistency)
The report /SAPAPO/SDRQCR21 checks sales order requirements and delivery
requirements in SAP R/3 / ERP and SAP APO/SCM.
Only this report checks precisely the SD order tables (mapping tables: /SAPAPO/posmapn,
/SAPAPO/ordadm_i, /SAPAPO/schedlin, /SAPAPO/sdqtvb,...) These mapping tables checks
are not performed by the CIF delta report!
It is recommended to use this report if the /SAPAPO/cif_deltareport3 report cannot correct
the inconsistencies, especially inconsistencies of specific incorrect mapping tables. If you are
using the backorder processing (BOP), especially with Rule Based ATP, you should use this
report regularly to check the consistency of the SD mapping tables, as BOP relies on these
tables being consistent. For more information, refer to SAP Note 444641.
If you run report SRDQCR21 regularly on SAP R/3 / ERP, you can set the flag “read
requirements from table VBBE” in report /SAPAPO/SDRQCR21 to improve its runtime
significantly (which is the commonly used procedure for checking a large volume of orders).
For few orders, it is quite practical to use the (much slower) option “build requirements from
document flow”.

Recent enhancements of /SAPAPO/SDRQCR21 (by SAP note 987299)


The redesign of report SDRQCR21 (see section 3.3.1.5) resulted in a complete new coding
compared to the previous version of SDRQCR21 with respect to regenerating the
requirements. In order to avoid redundancy, a new common interface was created and
/SAPAPO/SDRQCR21 will use the coding of the new version of SDRQCR21.
In addition, the following improvements were integrated into /SAPAPO/SDRQCR21
The possibility to check specific order numbers or order items
Iteration (quite similar to the iteration of the CIF comparison): The report first selects
orders in SCM, then in R/3 and compares the selected orders. A new check is
triggered in case of inconsistencies until the inconsistencies disappear or the number
of maximal iterations is reached. A waiting time [s] can be defined from one iteration
step to the next.
Table /SAPAPO/OBREF (document flow) was only checked for RBA subitems. With
SAP note 987299, the report can identify and correct an incorrect document flow
including the document flow from Stock Transfer Orders to the delivery document.
More accurate evaluation and correction of Product Allocation Assignments
It is now possible to exclude fully delivered orders and/or to check sales orders or
stock transfer orders only which improves the run time.
Redesign of the requirement recompilation
If you use the option to build requirements from the document flow, a very long
runtime may occur and the requirements are recreated without document blocks.
Therefore, inconsistencies may occur during the operation. With the new version of
the report (for which notes 998102 and 1023543 are prerequisite) a new logic for
recompiling requirements is introduced. The following new switches are available
which allow you to use the report during system operation.
o The first Flag 'Lock documents...' sets a document block if the requirements
from table VBBE are rewritten in the R/3 system. This prevents the document
from being changed at the same time.
o The second flag writes a planning file entry (net change per material/plant) in
the R/3 system if the requirement situation has changed. The next MRP reads
this and plans accordingly.

3.2.10.1 Handling of Temporary Differences


Within the new enhancements of report SDRQCR21 a better handling of temporary
inconsistencies was introduced which is described in detail in this section. When using the
old version, you need to execute appropriate steps manually.
35
© 2008 SAP AG
36
Best Practice: Data Consistency Check
When entering a logical system, first an RFC is triggered to ECC to check if the notes
998102 and 1023543/new version of SDRQCR21 is applied. If yes, specific selection
fields accept input, otherwise keep greyed out. If Flag 'Lock documents...' can be
entered, the new version of SDRQCR21 is active and used for reconciling and
updating the VBBE requirements.

Figure 3.16: Specific functions available with new version in ECC

Example settings for iteration:

Figure 3.17: Definition of Iteration Steps


Example for the iteration execution: An order was changed in R/3; the CIF-Queue still hangs
in APO Inbound. Report /SAPAPO/SDRQRCR21 would find an inconsistency and resend the
order (unnecessarily) to APO. The report waits x seconds (for example, x=2 as shown in
Figure 3.17: Definition of Iteration Steps) with the iteration activated. If the CIF queue was
updated meanwhile, this sales order is now consistent and will disappear from the result list
with the next iteration step. During the execution of the report, the iteration step and waiting
time are displayed.

Figure 3.18: Display of Execution Time and Iteration


The refresh is triggered to resend all data identified as inconsistent after the last iteration
step as a real inconsistency is assumed in this case. The information on how many
inconsistencies remained after each iteration step may be displayed on the result screen by
using the list display “Sort by document number”.

36
© 2008 SAP AG
37
Best Practice: Data Consistency Check
Figure 3.19: Result of /SAPAPO/CIF_DELTAREPORT3

3.2.10.2 Check frequency


You do not need to run the report /SAPAPO/CIF_DELTAREPORT3 for sales orders if report
/SAPAPO/SDRQCR21 is scheduled on a regular basis as this is already carried out by the
report /SAPAPO/SDRQCR21.

The report /SAPAPO/CIF_DELTAREPORT3 selects customer orders which are


contained in active and inactive integration models. You should run report
/SAPAPO/CIF_DELTAREPORT3 for sales orders from time to time, to perform a data
cleansing of inactive SCM data.

You can always run the report /SAPAPO/SDRQCR21 on demand, selecting as little data as
required, for example selecting affected sales order numbers only.

3.2.10.3 Important SAP Notes


Component SAP Note Description
SCM-APO-INT-SLS 987299 Report /SAPAPO/SDRQCR21: Iteration options
SCM-APO-INT-SLS 444641 Correction of incorrect sales order requirements
SD-BF-AC 998102 SDRQCR21: Enhancements to support the check
SD-BF-AC 1023543 Function Group V03T enhancements for sdrqcr21
Table 3.8: Important SAP Notes

3.3 Tools to Check Consistency within ECC


3.3.1 Tools for Processes Involving SD and LE
3.3.1.1 General Remarks for SD/LE Corrections
Most correction reports in SD and LE correct documents by reading the document
information, recalculating and saving the changed information on the database. Basically,
this corresponds to performing transaction VA02/VL02/VF02, but the following side effects
can occur:
Using updates based on VDATU, data could be written in a period other than the
original document date
Some data could be recalculated with the validity of the date (for example, credit
management data).
Both time shifts may affect results in BW and LIS and need to be considered when correcting
SD/LE transaction data. As long as only a few documents are changed, the time shifts are
negligible, but for larger amounts of data written into the current period a reconstitution of
statistical data should be performed.
All correction reports should be handled very carefully as an observed inconsistency
could be normal system behavior after certain actions like archiving.

Example: The document flow indicates a missing delivery. In this case, the question before
attempting to correct the situation is whether the delivery itself is missing or whether the
information in the document flow is incorrect. The document flow needs to be corrected in the
latter case, while in the first case a delivery needs to be recreated. A correction attempt of
the document flow in the first case will cause inconsistencies with potentially disastrous side
effects.
NONE of these correction reports should be planned on a regular basis in correction
mode! They should only be used to correct single faulty documents after root cause
analysis and may destroy data if used carelessly.

37
© 2008 SAP AG
38
Best Practice: Data Consistency Check
Example: The run of a report for correction of document flow followed by SDVBUK00 after
archiving of billing documents and deliveries will open ALL sales documents again thus
causing a production down situation.

3.3.1.2 Incorrect Document Flow


Within standard SD, the document flow is derived from two locations: The document flow
table VBFA which is mainly used to calculate the document flow from a preceding document
to a follow-up document and entries in the documents on item level (fields VGBEL, VGPOS
in tables VBAP, LIPS and VBRP) on which the display of document flow from follow-up
document to preceding document is based due to performance reasons.
Depending on whether a document is missing or a document flow entry is missing, several
reports exist to correct the data:
ZZKVBFA6: This report is intended to correct the document flow from order to
delivery. This report deletes document flow entries in case the document flow has not
been deleted when a delivery was deleted. The report is given in note 74513. Please
note that for archived deliveries, the document flow is needed. A similar report is
RVKVBFA1; for details see SAP note 4483
ZVBFADEL: Report for correcting the document flow from order to delivery. This
report will delete the document flow in case a temporary document number (begins
with $) has been saved. The report is given in SAP note 134939.
ZZVBFA_COMPLETE: Report for correcting the document flow from order to
delivery. This report is the opposite of the above mentioned reports. While the former
reports delete the document flow, this report will create missing document flow entries
if a delivery exists for an order. See SAP note 427059 for details.
WSCOR001: Report for correcting the document flow from delivery to LVS-TO. This
report will delete the document flow from the delivery to the TO if a document flow
exists from delivery to TO but the transfer order has been deleted. The report is
provided by SAP note 70310.
WSLIPS02: Report for correcting the document flow from delivery to LVS-TO. This
report will delete the document flow from delivery to TO if the delivery has been
deleted but the flow to the TO still exists (SAP note 28448).
ZTRVBFAC: Report for correction of document flow from delivery to transport. If a
transport has been created for the delivery but the transport is not shown in the
document flow, this can be corrected by SAP note 396253.
ZZVBFA01: Report used for correction of document flow from delivery to invoice. This
report creates VBFA-entries similar to report ZZVBFA_COMPLETE but this time from
delivery to invoice. See SAP note 38587 for details.
ZZVBFA02: Report used for correction of document flow from order to invoice. The
report given in SAP note 152051 is very similar to report ZZVBFA01 but this report
creates the document flow between order and invoice in case of order related billing.
ZZVBFA03: Report for correcting sales orders after purchase order deletion. If the
purchase order is deleted, the field VBFA-RFMNG should be set to zero. This report
which is given in SAP note 521416 can be used to display and/or correct documents
where this was not the case.

3.3.1.3 Inconsistent Document Status


When talking about the status of SD/LE-documents, we need to distinguish between two
different statuses:
The object status
The document status
While the document status contains relevant information about the document’s status
regarding the SD/LE business process, the object status contains information about the
operational status of the document. The document status is available on header and item
level and is stored in the SD tables VBUK (key field VBELN) und VBUP (key fields: VBELN
38
© 2008 SAP AG
39
Best Practice: Data Consistency Check
POSNR) associated with the individual document. The object status is stored in the general
status tables JEST/JSTO and is only linked to the SD-document via field OBJNR.
The main correction report for the document status is report SDVBUK00. This report should
always be used if a document status was determined incorrectly for a sales document. The
report can only correct the status if the display of the status in the sales document differs
from the status on the database as in both cases the same calculation rules are applied. If
the same status is displayed in the document and on the database, the report SDVBUK00
will not correct the document. In this case, it is necessary to find the cause for the status
determined if you think they are incorrect.
The best way to find out why a document status should change is to debug the
calculation of the document status in the sales document using breakpoints in
function modules RV_XVBUP_MAINTAIN for item status information and
RV_XVBUK_MAINTAIN for the header status. Here, you can see in detail on which
data the individual status information is calculated and how the (incorrect) result is
determined.
A common root cause for an incorrect overall processing status of a sales order is the
maintenance of a completion rule in the customizing of the order’s item category (transaction
VOV7). The solution would be to remove the incorrectly maintained completion rule from item
categories for which no subsequent sales documents will be created during the normal
business process. The completion rule should only be entered in item categories which are
used for inquiries, customer quotations or contracts. Sales document affected by this
incorrect customizing can only be corrected by changing the customizing of the item
category, transferring the item category to existing sales document (as in most customizing
changes existing sales documents are not updated) by means of report ZZERLRE which can
be found in SAP Note 79193. After correction of the customizing error the document status
can be fixed using SDVBUK00.
Report SDVBUK00 should definitely not be scheduled daily and for all documents. Instead, it
should only be started if required for the correction of an individual document after careful
investigation.
A similar report exists to correct document status information in deliveries and is called
RVDELSTA. The working model is identical to SDVBUK00 but considers the specific status
information available in deliveries.
So far, the document status has been updated but also inconsistencies with the object status
can arise. Several correction reports exist to rectify errors within the object status link in the
SD documents.
SD documents with object number in SD-Tables but no corresponding status object in table
JSTO: Commonly, these cases will result in error message BS001 when changing the
affected documents. These inconsistencies can be identified using standard report
SDSTATU1 or the generic consistency check report delivered by ST13. In both reports you
can list entries on header and item level where field OBJNR is filled but no corresponding
entry exists in table JSTO. The found inconsistencies can be corrected using report
SDSTATU2. This report creates entries in the status tables if an object exists in VBAK or
VBAP but not in the status tables. After creation of the corresponding entries, the report also
sets appropriate status information.
SD documents with assigned object number in status tables but missing link in SD tables:
Two reports have been delivered by SAP notes correcting this information on header and
item level. Report ZONRVBAK corrects documents having an incorrect status object (VBAK-
OBJNR initial). The report ZONRVBAK fills VBAK-OBJNR if entries in the status tables exist
but the connection got lost in table VBAK. The corresponding report on item level is called
ZONRVBAP. Both reports are given in note 162869 and enhanced in note 413555

3.3.1.4 Inconsistency Between Billing and FI


A common inconsistency due to incorrect enhancements in user exits is FI documents
without a corresponding SD billing document. To check and monitor these cases, report
ZF_MISS_SD can be used which is available via OSS note 904810. The report
39
© 2008 SAP AG
40
Best Practice: Data Consistency Check
ZF_MISS_SD checks for FI documents (table BKPF) with the reference activity VBRK.
Afterwards, table VBRK is checked for corresponding entries with document number VBELN
equal to BKPF-AWREF. If no corresponding entry is found, company code, fiscal year,
period and document number of the respective FI document as well as the number of the
missing billing document (=BKPF-AWREF) are listed.
In a second list, the report displays the summary local currency amounts from the affected FI
documents, according to company code, fiscal year, G/L and customer account.
If report ZF_MISS_SD has identified missing documents and you have made sure that the
missing billing documents have not been archived, perform a transactional correctness
analysis of the coding. In all cases identified in the past, incorrect coding in user exits and
enhancements has been identified as the root cause where an implicit or explicit commit was
performed destroying the logical unit of work. To understand the impact of the coding
mistakes and where to find these, it is essential to understand the interaction between SD
and FI in this process.
During the posting of billing documents, document number gaps can occur sporadically
although the buffering for the SD document number object RV_BELEG is switched off as a
document number for the billing document is assigned early in the posting process in the
dialog task (Function module RV_INVOICE_DOCUMENT_ADD). Once the document
number has been assigned, this is passed on to FI to set up the tables of the accounting
interface. If a severe error occurs (for example, a termination in update task), the update is
terminated prematurely and all sub processes - including number assignment - are rolled
back. But a roll back can only roll back data up to the last database commit. This means that
a rollback always goes back to the last COMMIT command of the application, therefore
either to the beginning or the last successful update. This means in turn that gaps in a
number range only occur when the system cannot roll back the data completely due to
intermediate database commits. Unfortunately, database commits are not only triggered by
the ABAP command COMMIT WORK but also by a number of other actions like sRFC-calls,
screen changes, and so on.
In a standard SAP R/3 system, exactly one COMMIT command is used during the billing
process. This in turn means that it is required to examine the process of the billing for
possible enhancements (user exits, modifications, and additional printing programs) by
means of a code review or by runtime analysis (ST05) of the billing runs including the
following programs:
billing document programs
programs for the accounting interface
accounting programs
Besides this detailed investigation, the following organizational measures can help to
pinpoint the root cause of such inconsistencies:

Do not transfer billing documents automatically to the accounting; instead use the
two step procedure. In this procedure you set a posting block in the customizing of
the billing document type and use transaction VFX3 to post the billing documents
into accounting after creation of the billing documents.

By separating the two steps, it is possible to determine whether the root cause is to be found
within SD or FI coding. As a side effect, it is ensured that billing documents exist for all
created accounting documents as the existence of a billing document is a precondition for
the creation of the accounting document in this procedure. In addition, you should document
missing document numbers in FI by report RFBNUM00 (SAP note 148757) and document
update terminations (transaction SM13).

40
© 2008 SAP AG
41
Best Practice: Data Consistency Check
3.3.1.5 Inconsistent SD Requirements

3.3.1.5.1 Introduction
Sales documents (quotation, sales order, scheduling agreement, …) and deliveries have
requirements as long as they are not completed (completely delivered/cancelled/rejected).
Those requirements are written, updated, and deleted accordingly in table VBBE. A detailed
explanation of the handling of requirements can be found in SAP note 547277. Too many,
too few, or simply incorrect sales documents or delivery requirements may trigger follow-on
errors in planning, procurement (production, purchase order) or document processing
(availability check).

3.3.1.5.2 Report SDRQCR21


Run report SDRQCR21 to ensure VBBE data is correct. If an SCM system is used, it is also
necessary to run /SAPAPO/CIF_DELTAREPORT3 in addition for sales orders setting the
flag “Use Table VBBE for Sales Order Comparison”. Make sure all sales data is consistent.
See SAP note 425825 for details. See also the chapter “Tools to check consistency between
ECC/SCM” within this document.
How to solve reoccurring requirement errors: The most frequent errors have to be found first
by monitoring and analyzing the spool lists. The most challenging step is to find a
reproducible example afterwards. Errors in standard can be corrected by SAP development;
errors in custom enhancements can be corrected by the responsible developers.
Results of report SDQRC21 can be monitored with the Business Process Monitoring in SAP
Solution Manager with monitoring object DCSDQCR after applying SAP note 1036106.

3.3.1.5.3 Restrictions of Report SDRQCR21


No parallel processing with a data transfer: Only execute the SDRQCR21 report with a
data transfer once you have made sure that no one will process sales documents or
deliveries (for the materials/plants according to the selection) while the report is running. The
reason for the restrictions is that the report does not read or set blocks. It therefore produces
correct results only if no other program updates sales or delivery requirements, or checks
availability while the report is running. The report SDRQCR21 may generate VBBE
inconsistencies (temporary or “fake” errors) when it runs in parallel to requirement relevant
transactions (for example, online transactions, backorder processing, mass delivery creation,
mass batch determination, cancellation of sales orders, IDocs, EDI). The item requirements
are wrongly calculated due to these parallel activities.
Further details can be found in SAP Note 25444 describing report SDRQCR21 in detail.

3.3.1.5.4 New Version of Report SDRQCR21


Large companies with global installations are often running 24 x 7 businesses with
high volumes that make it nearly impossible to find time windows with a low system
activity regarding sales orders and deliveries. Therefore, it was necessary to improve
the report SDRQCR21. For the above mentioned reasons, the report SDRQCR21 was
redesigned and an updated version was provided in SAP note 998102 (SD-BF-AC).
The new version of SDRQCR21e has a new enhanced selection screen:

41
© 2008 SAP AG
42
Best Practice: Data Consistency Check

Figure 3.9: New Selection Screen of SDRQR21


A correction is also possible by document number within a very short runtime and/or
document date. The other new flag is the Enqueue flag.
Regarding compare and non-compare mode: With locked mode active, the compare
mode is useful because you are locking only documents with inconsistencies in requirements
whereas the direct mode (non-compare mode) regenerates VBBE for every item (with or
without requirement errors) and therefore locks every document.

It is recommended to use the compare mode if you do not want to lock many
documents. This compare mode should be the normal usage of SDRQCR21.

Exceptional case: X is the percentage of items with requirement errors in your system
(number of items with requirement errors / total number of items selected). It becomes
interesting to switch to the non-compare mode instead of the compare mode when X is very
high because you skip the pre-selection phase and the compare routines to directly
regenerate the items requirement. This non-compare mode should be used exceptionally
(only in case the whole VBBE is corrupted for one or several material plant combinations).
Regarding the runtime of locked and non-locked mode: The non-locked mode is always
faster than the locked mode. You can really start to see the difference when X becomes high.
However, if X is small, the difference should not be very significant. Non-locked mode is not
100% secure but the probability of errors is a lot smaller than for the old version of
SDRQCR21. It can be used without problem in a system with low activity.
Regarding the impact of other jobs on the runtime: The runtime of SDRQCR21 might
increase due to the locking process if there are many items which could not be locked. This
is due to the retry logic. The probability of an item failing to get a lock increases with the ATP
activity of the system. Therefore, a slightly longer runtime might occur if transactions V_V2 or
VL10 are running in parallel to SDRQCR21. Because SDRQCR21 locks the documents it
can influence the runtime of other jobs like VL10 or V_V2. They wait for or retry the
documents that were locked by SDRQCR21. Experience shows that the runtime
degradations should not be very significant.
The new version of SDRQCR21 can be used both during the week or weekend. It does not
really matter and results do not depend anymore on the system activities.
It makes sense to use the compare mode and use the locked (flag "enqueue") and non-
locked mode for tests in productive system. The result list with or without simulation mode is
the same for simulation runs (see SAP note 998102 for additional information).

3.3.1.5.5 Case Study: Temporary Requirement Errors


What is a temporary or fake error? A temporary or fake error is an error that is not created by
the system but by the execution of report SDRQCR21 in update mode in parallel to high
peak system activity with sales orders/deliveries. Identified temporary errors get updated on
the database which results in real errors; those errors would get detected and repaired in the
next execution of the report, but at the same time SDRQCR21 could again generate new
errors.

42
© 2008 SAP AG
43
Best Practice: Data Consistency Check
Background: In this case, the old version of SDRQCR21 was running daily in update mode -
with a large number of errors in the spool lists, until the middle of month 10. Then, the
execution of SDRQCR21 was switched to daily in simulation mode and update mode only on
Sundays. As of month 12, the new version of report SDRQCR21 was scheduled daily in
simulation mode at the same time as SDRQCR21 to compare the spool lists of both report
versions. In the future, only the new version of SDRQCR21 will be scheduled.
The graph indicates the number of changes of Sales Orders and Deliveries measured during
5 weeks, confirming that Sundays are the days with the lowest business activity.

# S O a n d L F C h a n g es : M in im u m o n S u n d a ys

60000
# SO and LF changes

50000

40000

30000

20000

10000

0
26.08.2006 02.09.2006 09.09.2006 16.09.2006 23.09.2006 30.09.2006
da y

Figure 3.20: Document Changes by SDRQCR21

The next graphs display the number of changes during a Monday and a Sunday. However, it
was not possible to identify or setup time windows with zero activity on Sundays due to the
business requirements, but the Sundays were chosen as the only and best day to schedule
SDRQCR21 in update mode.

Figure 3.21: Document Changes by SDQRCR21 on a Monday

43
© 2008 SAP AG
44
Best Practice: Data Consistency Check

Figure 3.22: Document Changes by SDQRCR21 on a Sunday

The next comparison chart shows the number of errors found in the spool lists for the old and
new version of report SDRQCR21. Evaluating the spool lists, it was found that about 97% of
the errors detected by the old version of SDRQCR21 were temporary errors (due to
production operation peaks on sales orders and deliveries while the report was running).

requirement errors "VBBE"


35.000

30.000

25.000
# of errors

20.000

15.000

10.000

5.000

0
06

06

06

06

06

06

06

06

06

06

06

06
06

06
20

20

20

20

20

20

20

20

20
20

20

20

20

20

day
9.

9.

0.

0.

1.

1.

1.

2.
9.

0.

0.

1.

2.

2.
.1
.0

.0

.0

.1

.1

.1

.1

.1

.1

.1

.1

.1

.1
15

22

29

06

13

20

27

03

10

17

24

01

08

15

sdrqcr21: incl temp. errors z_sdrqcr21e: real errors

Figure 3.23: Comparison Between Old (sdrqcr21) and New (z_sdrqcr21e) Version of SDRQCR21
Note that from the middle of month 10 on, the old version of report SDRQCR21 was running
in simulation mode only, except for Sundays. Running it daily in update mode, the number of
errors in the spool lists would have raised further than in month 9 and the beginning of
month10.
Running the report SDRQCR21 in update mode during high peak system activity with sales
orders/deliveries, the found temporary errors get updated on the database which results in
real errors; those errors would get detected in the next execution of the report.
44
© 2008 SAP AG
45
Best Practice: Data Consistency Check
Therefore, to analyze requirement errors, it is very helpful to schedule report SDRQCR21 on
a regular basis. Otherwise, it is very difficult to find the real errors as they get lost in long
spool lists.

Only the new version of SDRQCR21 is able to filter out the temporary errors avoiding
the creation of fake errors.

3.3.1.5.6 Root Cause Analysis of Inconsistent Requirements


If you regularly find requirement errors in the spool lists of SDRQCR21, you should check the
last changes in the history of the affected documents.
Example: Start the display of the change documents either by displaying a sales order in
transaction VA03 and using the menu path “Environment Changes” or starting the report
RVSCD100 directly in transaction SE38. On the resulting selection screen, enter the
document number and item and select “Time of change”

Figure 3.24: Displaying Change Documents

Figure 3.25: Changes Within a Document

45
© 2008 SAP AG
46
Best Practice: Data Consistency Check

Figure 3.26: Detailed Display of Changes


In this example, a reason for rejection was set as the last change in the document, which
may be the root cause of a requirement error. The probability that the requirement error is
due to a customer modification or user exit is very high.

46
© 2008 SAP AG
47
Best Practice: Data Consistency Check

3.3.2 Tools for Processes Involving MM


There are several tools to identify and correct inconsistencies in Materials Management. The
tools for customer usage are transaction MB5K (report RM07KO01) and the reports
RM07MMFI and RM07MCHB. For special cases, the reports ZZWASTOR and ZTRAME are
available as well.
The above mentioned reports – except ZZWASTOR – are intended to identify
inconsistencies. The correction can be performed only by specially trained SAP consultants
using analysis and correction reports from SAP note 32236. These reports shouldn’t be used
by customers as the results could be misleading without a deep technical understanding of
the application.
SAP note 32236 provides a collection of check and correction reports used exclusively by
SAP Support to analyze and correct stock inconsistencies. In general, only the stock and
valuation tables are corrected, but not the document tables (material & FI document). The
target of the correction is a consistent system; it does not matter during this correction
whether the stocks are in-line with the physical reality. This is handled by a physical
inventory check which should be performed afterwards.

These reports also cover the special stocks used by the Industry Solution “Aerospace &
Defense” in the MRO process. Special reports are provided for stock inconsistencies in the
Industry Solution “Oil & Gas” in SAP notes 67261, 212707, 378731, and 447714.

The SAP note 32236 contains the analysis report MBFIRST which gives a first overview of
the system with important details, for example, valuation level, negative stocks allowed,
Material Ledger usage, serial number usage, WM usage, archived material and FI
documents, and so on.. This information is very important when starting the correction
process.

3.3.2.1 General Inconsistencies within Materials Management


The standard transaction MB5K (report RM07KO01) is able to determine inconsistencies
within Materials Management. The report should be planned at least once a week in a time
period where as few as possible activities are carried out and can issue the following error
messages:
011 Calculation number missing
045 Incorrect price
060 Negative stock not allowed
063 Material ledger record does not exist for current period
066 Material ledger currency record does not exist for current period
500 Actual quantity not equal to total of stocks
501 Actual quantity not equal to total of valuation segments
If the errors 060, 500 or 501 occur, make sure the reports from SAP note 32236 are in the
production system. Then open a customer message under the component MM-IM-GF-INC
and fill in the checklist from SAP note 76926. The further analysis and the corrections will be
carried out by the SAP Support.
If the error 060 is issued, the report found negative stocks for materials although this has not
been allowed. Furthermore the transaction compares the stock tables (MARD, MARC,
MCHB, MKOL, and so on) and the valuation tables (MBEW, EBEW, QBEW, OBEW) (error
500) respectively the valuation segments for split valuated materials (error 501).
In the case of errors 063 and 066, please contact the CO-PC-ACT support who will correct
the inconsistencies. You can also check the consistency between Material Ledger and
Materials Management with the transaction CKMC.
If the MB5K finds out that the calculation number is missing (MBEW-KALN1), you can
regenerate it directly from the MB5K.
The message 045 “Incorrect price” means that the current price in the material master
(MBEW-VERPR or MBEW-STPRS depending on the value in MBEW-VPRSV) differs from
47
© 2008 SAP AG
48
Best Practice: Data Consistency Check
the calculated price (total value of stock/total valuated stock – MBEW-SALK3/MBEW-
LBKUM). These kinds of errors can be corrected with a simple price change in MR21.
However, this is only the correction but it doesn’t solve the problem. Based on our
experience this error occurs mostly due to the following situation: the material has a quite low
price, for example, 50 EUR/1000 PC and only small quantities are moved, for example 10
PC. Due to the rounding, for example, at Goods Issues, the price itself doesn’t change but
the calculated price will slowly differ from the moving average price or standard price. To
solve the problem, the material master has to be changed, for example, the price unit should
be changed.

The monitoring objects DCMMMB5K and DCMLCKMC exist to facilitate the regular
monitoring of transaction MB5K and CKMC using the Business Process Monitoring
functionality of the SAP Solution Manager.

3.3.2.2 Inconsistencies of Split Valuated Materials within MM


The report RM07MCHB helps to determine inconsistencies for split valuated materials. It
checks the mother segment against the daughter segments, for example, whether the
quantity of the mother segment and the sum of the daughter segments’ quantity are the
same.
If the report shows errors, please make sure the reports from SAP note 32236 are in the
production system. Then, open a customer message on the component MM-IM-GF-INC and
fill in the checklist from SAP note 76926. The further analysis and the corrections will be
carried out by the SAP Support.

3.3.2.2.1 “Dead” Stock-in-Transit


The report ZTRAME can be used to identify “dead” stock-in-transit that can be removed by
normal bookings. The full details about the report can be found in the SAP note 392205
where the report itself can be found, as well. The note also describes how to use the
movement types 557/558.

3.3.2.3 Inconsistencies Due to Doubled Goods Issues for Deliveries


The report ZZWASTOR is introduced in SAP note 424414. The purpose of this report is to
cancel material documents booked via delivery note by mistake twice. It simply cancels the
material document that is not needed without updating the delivery note which is not possible
by using the standard transaction VL09.

3.3.2.4 Inconsistencies (quantity) Between Material Documents and Stock


Tables
The report MBSTOCK compares the material documents and the stock tables. During the
run the table MSEG is read and the should-be stock is calculated and compared with the
relevant stock table fields. You can find a list of stock tables later in this document. The
report presumes that the stock calculated from the material documents is correct and
therefore always suggests a correction in the stock tables. It is very important to know
whether there are archived material documents as this influences the report. If there are
archived documents, the results must be handled with great care! Therefore, the results
should only be analyzed by experienced SAP consultants.

As the report reads the table MSEG, the run can be very performance intensive if the
selection is not specific enough.

48
© 2008 SAP AG
49
Best Practice: Data Consistency Check
3.3.2.5 Inconsistencies (quantity) Between Valuation Tables and Stock Tables
The report MBQUANT compares the stock tables against the valuation tables. The list of the
valuation tables can be found later in this document. The performance is not an issue in most
of the cases. This report assumes that the stock tables have the correct information over the
valuation tables thus it suggests a correction there.

It is very important to check the consistency between MM and FI before correcting


with MBQUANT: If MM and FI are consistent and table MBEW is corrected with report
MBQUANT, a MM-FI inconsistency is created.

Archived documents have no influence on this report. For very big selections MBMSSQUA
has to be used.

3.3.2.6 Inconsistent Stock-in-Transit


MBTRAME deals with real stock-in-transits from intra-company stock transport orders. It
checks & corrects the content of the field MARC-TRAME.

3.3.2.7 Purchasing
The report ZCORREKESBYGR corrects the quantities reduced (MRP) (EKES-DABMG) of
confirmations with GR assignment. Further information can be found in SAP note 215072.
The report RM06C020 finds and corrects inconsistencies with dependent requirements for
subcontracting purchase order proposals. Further information can be found in SAP note
115899.
With the correction report ZKORVETV, the table VETVG (Delivery DueIndex) can be set up
for an individual purchase order again. Further information can be found in SAP note 61148.
SAP note 100690 provides correction reports ZKO* for inconsistencies of stock transport
orders and stock transport scheduling agreements regarding quantities at schedule line level
(EKET-GLMNG, EKET-WAMNG, EKET-WEMNG).
SAP note 202875 provides the correction report ZCORRDABMG regarding inconsistencies
of EKET-DABMG and EKES-DABMG.
The report ZKOREKET checks and corrects if EKPO-MENGE of a scheduling agreement
does not correspond to EKET-MENGE. Further information can be found in SAP note 67278.

3.3.2.8 Invoice Verification


The report ZREPMIR7 finds and corrects invoices without FI follow-up document or entry in
the purchase order history. Further information can be found in SAP note 382797.
The report ZREP_MIRO_REMNG finds purchase order items with negative total invoiced
quantity (REMNG). The correction can only be done by SAP. Further information can be
found in SAP note 491074.
The report Z_FIND_DOUBLE_FI finds logistics invoices with incorrectly created duplicated FI
documents. A correction of this issue can only be done by SAP. Further information can be
found in SAP note 674190.

3.3.3 Inconsistencies between MM and FI


3.3.3.1 Inconsistencies between MM and FI (Valuation Tables vs. FI Stock
Accounts)
Report RM07MMFI can identify inconsistencies between Materials Management and
Financial Accounting. The report compares the valuation tables (MBEW, EBEW, QBEW,
OBEW) directly with the stock account balances from table GLT0 with the help of the account
determination (table T030). The report should be used instead of the transaction MB5L
(report RM07MBST) as it is much less performance intensive. The report should be started
during a timeframe where no bookings take place to get a stable result. The report should be
49
© 2008 SAP AG
50
Best Practice: Data Consistency Check
run 3-5 times to eliminate errors resulting from bookings during the analysis if the report
cannot be executed during such a time window.
Make sure the reports from SAP note 32236 are in the production system if the report
shows stable results and has identified inconsistencies. Open a customer message
on the component MM-IM-GF-INC afterwards and fill in the checklist from SAP note
76926. The further analysis and the corrections will be carried out by the SAP
Support.

The Business Process Monitoring object DCMFIDIF is available to facilitate the


regular monitoring of inconsistencies.

The reports provided by SAP note 32236 should only be carried out by experienced SAP
consultants as the results of these reports need some interpretation. SAP note 34440
describes the correction procedure applied by the SAP consultants for customers.

3.3.3.2 List of Relevant Stock and Valuation Tables


Below, you can find a list of the relevant valuation and stock tables. The IS-ADEC tables are
mentioned as they are covered by SAP note 32236, the special IS-OIL tables are not, as
they are handled separately by special reports. The tables ending with H are the history
tables. Further information on history tables can be found in SAP notes 171465 and 193554.
History tables should be corrected only in exceptional cases. For details see SAP note
34440.
Important valuation tables:
MBEW Material Valuation
MBEWH
EBEW Sales Order Stock Valuation
EBEWH
QBEW Project Stock Valuation
QBEWH
OBEW Valuated Stock with Subcontractor
OBEWH

Important stock tables:


MARC Plant Data for Material
MARCH
MARD Storage Location Data for Material
MARDH
MSSA Orders on hand Total
MSSAH
MSKA Sales Order Stock
MSKAH
MKOL Special stock from vendor (vendor consignment, RTP)
MKOLH
MSSL Subcontracting stock at vendor Total
MSLB Subcontracting stock at vendor
MSLBH
MSSQ Project Stock Total
MSSQH
MSPR Project Stock
MSPRH
MSKU Special stock at customer (customer consignment, RTP)
MSKUH

50
© 2008 SAP AG
51
Best Practice: Data Consistency Check
Important stock tables available only for Discrete Industries
MCSD Customer Stock
MCSDH
MCSS Total Customer Stock
MCSSH
MSCD Customer stock with vendor
MSCDH
MSCS Customer stock with vendor - Total
MSFD Sales Order Stock with Vendor
MSFDH
MSFS Sales Order Stock with Vendor - Total
MSID Vendor Stock with Vendor
MSIDH
MSIS Vendor Stock with Vendor – Total
MSRD Project Stock with Vendor
MSRDH
MSRS Project Stock with Vendor – Total
3.3.4 Tools for Processes Involving WM
3.3.4.1 Stock Comparison Between WM and IM
Transaction LX23 investigates and corrects stock differences between inventory
management and warehouse management. LX23 can be used both for a centralized and for
a decentralized scenario. LX23 runs on the warehouse management system in the case of a
decentralized scenario and the inventory management stocks are determined by RFC from
the central ERP system. The actual differences between the two inventory management
levels can be compared via an automatic physical inventory where the stock comparison
report first reads all IM stocks including special stocks and reads and summarizes the WM
stocks in a second step. The individual stocks are listed and the difference is calculated.
An automated correction can be executed if you select the 'Clear differences' check box, but
you should bear in mind that the system always assumes that the WM system is the leading
system. Correspondingly, in the event of differences, the IM stock is adjusted to the WM
stock and NOT vice versa. This correction is performed by a generated batch input session
for transaction MI10 which performs an automatic physical inventory which adjusts the IM
stocks to the WM stocks. No batch input session is generated in the case of a decentral
warehouse scenario, but IDocs of message type MBGMCR are created instead and sent to
the central ERP system.

3.3.4.1.1 Restrictions of Transaction LX23


Active handling unit management: Transaction LX23 cannot be used for correction of HU-
managed storage locations. However, for storage locations with HU management, you can
use transaction LX23 to display stock differences between WM and IM. HU managed stocks
cannot be analyzed at all. SAP note 660009 describes how to proceed for HU managed
storage locations in case of stock differences for one of the three levels IM, WM and HU.
Materials with serial number requirement: Clearing the differences via a batch input
session which has been created by LX23 is not possible for materials with serial number
requirement. More information and how to proceed in this case can be found in SAP note
484788.

3.3.4.1.2 How to Handle Temporary Differences with Transaction LX23


You have to keep in mind that temporary stock inconsistencies between warehouse
management and inventory management, will always exist in the case of a decentralized
scenario when analyzing stock differences due to the asynchronous update between the
systems. Before running LX23 in a decentralized scenario you need to ensure that all IDocs
51
© 2008 SAP AG
52
Best Practice: Data Consistency Check
of message type SHP_OBDLV_CONFIRM_DECENTRAL,
SHP_IBDLV_CONFIRM_DECENTRAL and MBGMCR are transferred and processed
successfully from the decentralized warehouse management system to the central ERP
system. Moreover, ensure for both cases - running LX23 for a centralized and for a
decentralized scenario - that stock changing postings take place neither on IM level nor on
WM level.
The report RLTREORG2 detects inconsistencies either between the quants and the storage
bins or between the quants and the storage units, assuming that the quant data is correct.
Further information can be found in SAP note 607226.

3.3.4.2 Correction of Picking and WM Status in Deliveries


The report RVDELSTA checks whether the status in a delivery is incorrect according to
related system information and must be determined again. This is also done for the picking
and WM status. More information can be found in SAP note 506510.
3.3.5 Tools for Processes Involving PP
3.3.5.1 Introduction:
If pure R/3 production planning is used (no other system keeping the data redundant)
external consistency problems are not an issue. In general, it is possible to record
confirmations with legacy systems for production control. The PP module allows downloading
operations of orders into external systems. Beforehand, the work centers have to be
transferred to the external system as well. The operations are then confirmed in the legacy
system and the confirmations are uploaded into PP subsequently. Upload and download are
performed with the transactions CI42 to CI45 within PP. The following recommendations only
refer to PP and do not consider if an SCM system or other external systems are integrated to
R/3 PP.

3.3.5.2 Erroneous goods movements


The stock situation is not updated whenever goods movements fail. Following requirements,
calculations (in R/3 or SCM) will work with incorrect stock levels and might therefore not
generate necessary receipt elements or unnecessary receipt elements. This may lead to a
production standstill if necessary materials are not available and ATP checks in productions
and sales orders will produce wrong results. The actual costs that are calculated for the
production order are not correct if the goods movements to the production order are not
booked correctly. The costs of production are in such a case not correctly reflected, which
might have severe business consequences. Incomplete cost postings result in the end in
customers paying not enough.
To reduce such impacts and as a preparation for a reliable period end closing, the following
reports should be carried out regularly as standard activities:

3.3.5.2.1 Failed Goods Movements in PP-SFC


Report CORUAFFW/CORUAFWP (transaction COGI) exists for identifying and correcting the
reason for erroneous goods movements and subsequent post processing. The
stock/requirements situation in APO is updated accordingly! Compared to report
CORUAFFW, CORUAFWP is predestined for batch processing having interesting additional
features for batch processing as parallel processing, parameters for maximal number of
locks, maximal number of document positions, and so on.

The regular monitoring of failed goods movements can be facilitated with the
Business Process Monitoring in SAP Solution Manager using monitoring object
KPCONF01.

52
© 2008 SAP AG
53
Best Practice: Data Consistency Check
3.3.5.2.2 Failed Cost Postings in PP-SFC
Report CORUCOFC (transaction COFC) is used for identifying and correcting the reason of
failed cost postings and subsequent post processing. See SAP note 565370 “FAQ:
Reprocessing incorrect actual costs (COFC)” for details. Such costs are only of interest for
the R/3 system and are not relevant for SCM nor do they have an impact on APO PP/DS.

3.3.5.2.3 Failed Goods Movements from Repetitive Manufacturing


Reprocessing of failed records from repetitive manufacturing may be processed online with
COGI or in the background using either the report RMSERI13 or COGI with batch input. The
reports CORUAFWP and CORUPROC are not working for repetitive manufacturing. The
tools cannot be clearly classified as consistency checks or housekeeping jobs. For
completeness, the reports when using decoupling are mentioned:
Postprocessing of goods movements in PP-SFC (only relevant when decoupling is
activated): report CORUPROC (transaction CO1P)
Postprocessing of Cost in PP-SFC (only relevant when decoupling is activated):
report CORUPROC (transaction CO1P)
Both jobs should be regular business steps when using decoupling and are not related to
failed postings.

3.3.5.3 Important Notes for PP Reprocessing:


Component SAP Note Description
PP-SFC-EXE-CON 565370 FAQ: Reprocessing incorrect actual costs (COFC)
PP-SFC-EXE-GM 544520 FAQ: Picking
PP-SFC-EXE-GM 540392 FAQ: Automatic goods movements
MM-IM-RS-RS 519580 FAQ: Reservations
Table 3.10: Important Notes for PP Processing

3.3.6 Tools for Processes Involving PS


3.3.6.1 Introduction
Usually, only a pure R/3 project system is used. As a result, there are generally no external
consistency problems, as no other system keeps the data redundantly. Therefore, a best
practice for consistency check in PS is not available. Nevertheless, some generic system
internal checks should be carried out under special circumstances. On the application side,
networks and orders, especially in big projects, should get charged, closed and then where
applicable get archived. As preparation for a reliable period end closing, the actions
described in the following sections should be carried out to ensure a consistent state.

3.3.6.2 Failed Postings from Confirmations

3.3.6.2.1 Failed Automatic Goods Movements


As a general impact, erroneous goods movements that are not post processed may lead to a
mismatch of the stock/requirements situation between system and reality which might falsify
ATP checks and any requirements planning for material components. In order to resolve
these issues, transaction COGI (report CORUAFWP) should be used regularly for post
processing of erroneous goods movements. (This also applies to PP and PM).

The regular monitoring of failed goods movements can be facilitated with the
Business Process Monitoring in SAP Solution Manager using monitoring object
KPCONF01.

53
© 2008 SAP AG
54
Best Practice: Data Consistency Check
3.3.6.2.2 Failed Postings of Costs
Impact of erroneous cost postings: Erroneous cost postings lead to missing costing records,
which for instance could be missing for resource related billing. That would cause the end
customer to be billed less. You should run transaction COFC (report CORUCOFC) for post
processing of erroneous cost calculations on a regular basis to decrease the impact of such
situations. This also applies to PP and PM.

3.3.6.2.3 Failed Non-Interactive Confirmations


Non-interactive confirmations can occur from the transfer of postings from the cross
application time sheet (CATS) or from posting confirmations via the PDC data interface -
communication channel KK4. Consequently, the actual hours, costs, and automatic goods
movements are not posted with all consequences for subsequent processing like incorrect
stock, missing costs, too little billing, too little work in progress, and so on.
To decrease the impact of such failed confirmations, you should run transaction CN30
(CORUPROC) for post processing of the error pools for non interactive confirmations on a
regular basis. This also applies to PP - transaction CO10N and PM - transaction IW46.
These three steps are also described in the SAP note 570471.

3.3.6.3 Value Category Customizing


The values for costs/revenues, payments/receipts are not reliable anymore in the Hierarchy
Reports (for example, transaction CJE0), the Structure Overview (transaction CNS41 /
CN41) and the Project Planning Board (transaction CJ2B) if the customizing of the value
categories is incorrect. Strange effects like disappearing costs/revenues, mixing of
costs/revenues, or costs appearing as revenues are possible without having this customizing
set up completely and correctly. The completeness of the value category customizing should
be checked in consistency check transaction CJVC according to SAP note 676123. Please
adjust your customizing following the output of the consistency check. After having corrected
everything, the project information database must be rebuilt by transaction CJEN (only, if
really something has been adjusted in this customizing).

3.3.6.3.1 Special Topics:


Correction reports for wrong commitments in PS reports are provided by SAP note 152571.
A general FAQ regarding inconsistencies in the area of the Project System is provided in
SAP note 678522.

3.3.6.4 Overview of SAP Notes for PS Consistency:


Component SAP Note Description
PS-CON 570471 FAQ 2: Confirmations in the Project System
PS-IS-ACC-HIE 676123 Project Information Database / Value Category
Customizing
PS-IS-ACC-HIE 678522 FAQ 2: Data inconsistencies in the Project System
Table 3.11: SAP Notes for PS Consistency

3.3.7 Tools to Check the Logistics Information System


Report RMCSUTIC provided by SAP note 382977 can be used to verify the customizing as
well as the definition of the information structure for customer made info structures. As not all
of the control table entries delivered within the SAP standard are needed for an error-free
update, the system displays errors that are not errors in the case of a check of the update
rules for SAP standard infostructures S000 to S500.

54
© 2008 SAP AG
55
Best Practice: Data Consistency Check

3.4 Tools not Depending on a Particular System


As soon as more than one system is involved, the consistency check has to cover the
interface related parts. In the SAP environment ALE (IDocs) are frequently used to transmit
data between systems, and here the following tools are relevant.
On the sender side, the following tools/checks have to be done: Are the interfacing
documents already out of the system, or are they still in outbound processing?
With transaction BD87, you can check the status of IDocs and the tRFC communication
status. Make sure that the interface data has completely left your sending system. For all
other interface technologies you have to use the available tools that the interface and used
middleware offer you. If you are using the SAP Exchange Infrastructure as your interface
middleware, you have to check the status of your XI messages in the XI Adapter Framework
as well as on the XI Integration Server for complete processing (for example, using the
Runtime Workbench – Message Monitoring). If you are for example transferring your
interface data via files, you have to make sure that all files are transferred completely.
On the receiver side, you have to check that the relevant interface data has been processed
completely and correctly. As long as data is still in status to be processed or in error status
on the receiver side, the executed consistency check report will not provide correct data. For
SAP systems with incoming IDocs, you use the same transaction BD87 to check the status
of your inbound IDocs. For SAP systems with incoming XI messages via ABAP Proxy
communication you use transaction SXMB_MONI to check the processing status of inbound
messages. For non-SAP systems you have to check for tools and processes, how to make
sure that all relevant interface data for the processing area has been completely and
successfully processed.
3.4.1 Tools for ALE Consistency Check
If you want to check the complete status of your transferred IDocs in the sending system, you
have to setup ALE Audit and ALE tracing.
For more information on this topic, please check the documentation in SAP Online Help
under:
http://help.sap.com/saphelp_erp2005vp/helpdata/en/0b/2a655d507d11d18ee90000e8366fc2/
frameset.htm
3.4.2 Tools for ALE Recovery
In distributed environments, you can also recover an SAP system following a system failure
caused by a database error. Using backup and point-in-time recovery, you can reset the
database to the status at the time the system failed and continue working with it.
In this case, transactions may still have taken place after the time the system is reset to,
which are subsequently lost and have to be carried out again. External interactions (for
example, ALE, EDI, mail, fax, telex), possibly carried out under a different key, are duplicated
and require special processing. For example, if a process sent a message to another system
and started another process here during the time between when the error occurred and when
it was discovered on the recovered system, the data related to this process will no longer
exist on the recovered system. The results of the second process carried out on the receiving
system still exist in this system. The inconsistent data in the two systems can be put right by
canceling the data resulting from the communication in the receiving system.
In point-in-time recoveries, certain documents and messages must be canceled in the
communication partners’ systems and messages sent earlier from the receiving system must
be sent again. You have to determine all the actions that need to be carried out in the
recovery process. The ALE Recovery tools help you to do this.
For Tools and more see SAP Online Help and topic “ALE Recovery for Data Consistency”
under:
http://help.sap.com/saphelp_erp2005vp/helpdata/en/26/29d829213e11d2a5710060087832f8
/frameset.htm

55
© 2008 SAP AG
56
Best Practice: Data Consistency Check

3.4.3 Generic Consistency Check for Two Linked Tables in One System
3.4.3.1 Introduction
Very often, you have to verify data between two different tables which are linked. For
example you could be interested in identifying line items without corresponding header
information (for example, sales order items VBAP without corresponding sales order header
data in VBAK). Or during data load/creation into a new system all materials of a certain
material type (for example, HAWA) should have a specific entry in one of the material master
views (for example, MARC or MVKE). Usually, you would have to write a program or a query
to find this kind of data. Within the service software portal transaction ST13, a new report
/SSA/Z>CD1 has been delivered that can answer all these questions.
The report starts with a selection screen where you can enter the table names you want to
investigate as well as field names and restrictions. You can also differentiate what kind of
check you would like to execute between the linked table (e.g. whether you would like to see
entries where a field has an undesired content or whether you would like to identify missing
fields).
The report creates an appropriate dynamic SQL-statement for the entered table names, use
case, and field restrictions. Once the selection has finished the resulting list is displayed
together with the used SQL-statement in the ALV. As the created SELECT-statement may
not be supported by an appropriate index, the run time of the report may be very long in
certain cases.

The results of regular executions of this report may be monitored from SAP Solution
Manager with the monitoring object “DCGEN001”.

3.4.3.2 Use Case 1: Completely Missing Parent Table Entries


This scenario of the report provides the possibility to list all entries of a dependent table
where no corresponding entries exist in a leading table. A typical usage scenario would be
line items without corresponding header information (for example, sales items in table VBAP
without header information in table VBAK). This information could be of interest after a data
base crash or after a report accidentally has deleted some data. In report /SSA/Z>CD1, you
need to enter the leading table as “Table 2” and the dependent table in field “Table 1”. In
addition, you need to specify the foreign key relation between the two tables in the area
“Define join condition between tables”. This case is activated by setting the flag “missing
entries in table2”. A typical example can be found in Figure 3.27: Finding Missing Header
Information where the range is restricted to certain document numbers using a select-option.

56
© 2008 SAP AG
57
Best Practice: Data Consistency Check

Figure 3.27: Finding Missing Header Information


After starting, the report creates an appropriate SQL-statement for the entered table names,
use case, and field restrictions and displays the resulting list in the ALV (see Figure 3.28:
Result Screen of the Generic Report). The upper part of the screen lists the total number of
hits (inconsistencies) found while the middle part contains further details which can be used
for detailed investigations or correction. The used SQL-statement is displayed in the lower
part of the screen. In the example we see that we have an item of sales order 47870 in the
database while no header information exists for this document number.

57
© 2008 SAP AG
58
Best Practice: Data Consistency Check

Figure 3.28: Result Screen of the Generic Report


As the created SELECT-statement may not be supported by an appropriate index, the run
time of the report may be very long in certain cases.

3.4.3.3 Use Case 2: Incorrect Field Content


While the first option has shown table entries where no corresponding information was found
at all in a leading table, sometimes it is also of interest whether two tables contain a correct
field content. A typical example would be to verify the quality of master data after a cut-over
or to check for outdated information after a customizing change. Here, the leading criteria
may be contained in another table than the actual field containing the value (for example, do
all materials of type HAWA (table MARA) have the intended MRP type in table MARC?). In
order to use the report for this use case, both tables need to be specified. In addition, the link
between the tables and all selection criteria that define the condition to be checked have to
be entered. In the example of comparing MARC and MARA this would be the material
number (MATNR) as a link between the two tables in the screen area “Define join condition
tables” and the criteria MTART = HAWA to restrict to the desired material types and the
MRP-type to be used by this material (see Figure 3.29: Example to Find Table Entries not
Having the Desired Field Content). The other criteria are only entered to restrict the hit list in
the given example.

58
© 2008 SAP AG
59
Best Practice: Data Consistency Check

Figure 3.29: Example to Find Table Entries not Having the Desired Field Content
While we have checked so far individual field content with a specified field content, you may
very often also interested in a consistency check between field contents of two linked tables
independent of the value. An example would be verification between customizing settings
and the associated stored values in the movement data. In this case, the only required
information is whether the values correspond, but not actually which field content it is
supposed to be. This type of check can be performed with the report as well. An example for
a selection variant is displayed in Figure 3.30: Compare Two Linked Tables without Knowing
the Intended Field Content.

59
© 2008 SAP AG
60
Best Practice: Data Consistency Check

Figure 3.30: Compare Two Linked Tables without Knowing the Intended Field Content

3.4.4 The Generic External Compare


Another generic tool to verify data consistency where no standard report is available is the “Generic
External Compare”. This tool set facilitates the extraction of data from two systems and the following
comparison. For details see the document “Documentation for the Generic External Compare”
available in the service marketplace under http://service.sap.com/bpm.

60
© 2008 SAP AG
61
Best Practice: Data Consistency Check

4. Appendix
4.1 General Roadmap for Analysis
Customer reports
difference

Perform Inconsistency
Assessment

Verify low-level data

Inconsistencies in
Y low-level data? N

Identify logically connected


steps

Inconsistency
with new data?
Compare connected steps
using check reports/queries
starting from bottom up

N
Y N
Inconsistency
in step data?
Trace and debug
underlying programs to
find RC
Y

N Intermediate
technical
steps?

Y
Trace and debug step, Trace and debug intermediate
investigate logical correctness steps

Correct Root Cause

61
© 2008 SAP AG
62
Best Practice: Data Consistency Check

Record and Evaluate


Business Process

Record and Evaluate Severe Detailed investigation


Inconsistency Impact Impact? whether productive
Usage is possible.

Record and Evaluate


Inconsistency Continue
Determination Process Productive
Usage

Inconsistenci Most Likely Performance


es Differences Optimization
disappearing
?

Several Verify
Systems Leading
involved? Systems

Questions for
Interface
Monitoring
New System
used?

Questions for
Transactional
Correctness

Questions
regarding Training Interface
Management
/ SMO SA

Verify Usage with


End Users

Detailed Analysis

62
© 2008 SAP AG
63
Best Practice: Data Consistency Check

Is a leading Reload data


system for from leading
the data system
available?

Does Is a working
dependent Backup
data exist? available?
Dependent

Is it sufficient How many Perform a


to reconstruct BO-types are point in time
data? corrupted? recovery

Recover data Do correction How many


by reports and recovery BOs are
reports exist? affected?

Write How many Perform a


recovery BOs are restore in parallel
reports affected? system

Perform a
manual
correction

Correct
dependent data

63
© 2008 SAP AG
64
Best Practice: Data Consistency Check

4.2 Dos and don’ts for Data Consistency


This section provides the most important data consistency rules for the areas monitoring and
error handling, program design, and end user training.
Monitoring for inconsistencies should always include:
Interface Monitoring
Business Process Monitoring
Update Termination Monitoring
Error correction and investigation should:
Be performed soon after error detection
Consider dependencies and data actualization before updating/reprocessing data
Perform corrections only if legally allowed
Never run correction reports for all documents without prior analysis
Ensure that no object changes are executed while running correction/check reports
Not perform data cleansing for temporary inconsistencies
Document number range gaps for later reference
Bear in mind that not everything possible is also allowed
Programs and enhancements should:
Follow the all-or-nothing principle
Consider implicit and explicit commits
Lock data to be changed using the SAP enqueue feature
Provide restartability features
Provide logging information
To prevent inconsistencies by data changes and to facilitate root cause analysis:
Customizing in the productive client should not be changeable
Change the maintenance status for tables that need to be maintained regularly only
for exceptional cases
Change management for customizing should be used
Customizing for existing data should never be changed. If you do so, consider the
impact on existing documents and use appropriate correction reports.
Changes of field contents using the debugger should be disallowed
Transports should never be made into a running system
End users should
Be trained in the handling of exceptions
Perform business process steps (and their physical pendant) completely and in the
correct order, for example, period end closing and period end closing reconciliation
Be aware of dependencies
Incorporate a data model of dependencies in the training
Business processes should:
Have a fall-back strategy in case of system and application down situations
Be designed rating logical correctness and clarity above possibilities of system usage
Be designed with a clear leading system having unique dependencies for each
business object

4.3 Advantages and Disadvantages of Different Recovery


Methods
There are different possibilities to recover data. This appendix lists advantages and
disadvantages of individual data recovery methods and provides guidelines for deciding
which method should be considered first. Please bear in mind that even though one recovery
method may seem best from an issue point of view (complexity, quantity), another one may
be the one to be used, as under certain circumstances the requirements for the preferable
method may not be met.

64
© 2008 SAP AG
65
Best Practice: Data Consistency Check
No Recovery Method Advantages Disadvantages Requirements

1 Complete restore of Consistent state of In distributed Time of first


backup (point in time complete system systems only occurrence of
recovery) Pure technical manageable failure needs
procedure to with high effort to be known
correct complex Data after Correct and
incosistencies “point-in-time” consistent
needs to be backup needs
reconstructed to be vailable
(for example, Persistance
from printouts layer must be
and so on) able to perform
During restore a point-in-time
and for some recovery
follow-up
activities no
productive
usage of
system
possible
2 Restore in a sandbox- Recovery of Data after Same
system and load of data individual tables “point-in-time” requirements
by RFC and table entries needs to be as “complete
possible reconstructed restor of
A productive (for example, backup” plus
usage of system is from printouts second system
in general possible and so on) with enough
Programs to ressources for
load data processing
between and storage
systems need
to be
developed
individually
3 Reload data from a A productive Dependent Leading
leading system usage of system is data needs to system
in general possible be containing
Simple technical reconstructed correct data
method utilizing from loaded
programs which data
are usually Usually only
available already for master data
available
4 Recovery from dependent A productive Often, Relatioship
data and systems usage of system is correction between data
in general possible reports need to to be
Many tools and be developed recovered and
reports available individually underlying
within SAP data
standard
environment
(DIMA, CIF-
Reports, LX23,…)
5 Manual correction of data Fast correction of Manageable
(debug mode, SE16, single corrupted quantity and
application transactions) data complexity
A productive No existing
65
© 2008 SAP AG
66
Best Practice: Data Consistency Check
usage of system is correction
possible reports
Table 4.1: Comparison of Recovery Methods
Taking into account the general pros and cons of the recovery methods, the deciding factors
for the individual methods are:
Methods 1 and 2 are the choice if data is contained in another system but is also
affected there, for example, the dependent data is already corrupted or does not exist
at all.
Method 1 should only be attempted if the data cannot be reconstructed by other
means because too many business objects have been destroyed and no
relationships usable for reconstruction exist.
Method 2 should be preferred over method 1 if the data cannot be restored by
relationships or dependent systems but only a few tables or business objects are
affected (in contrast to a widespread failure where method 1 should be preferred). It
is advisable when the complexity of the data to be restored is manageable.
Methods 3 and 4 should be preferred if data is contained in another system or can be
reconstructed from further existing data as both allow productive usage of the system
in most cases.
Method 5 should only be attempted if only a very small amount of data is affected and
no complex relationships exist.
When the issue is very complex, meaning that several objects with extended
relationships are affected, a reconstruction may logically be impossible or too time
consuming. In this case, methods 1, 2 or 3 are optimal.
Method 1 is not applicable if the time between error detection and first occurrence
(point-in-time for which data needs to be restored) is very long as too much data
needs to be reconstructed from external sources like printouts, memorization, and so
on.
If we look at these observations and recommendations, the following picture emerges with
respect to quantity and complexity:

66
© 2008 SAP AG
67
Best Practice: Data Consistency Check

Quantity

3
4

Complexity

Figure 4.1: Quantity and Complexity of Different Recovery Methods

4.4 Further Information


4.4.1 Related information
Important information regarding monitoring of data consistency with the SAP Solution
Manager can be found in the document “Setup Guide – Data Consistency Monitoring”
available in the Media Library under http://service.sap.com/bpm.
Documentation of a general tool to compare data in different systems with the help of SAP
Solution Manager can be found there as well.
If you need more information regarding Continuity Concepts refer to the documents
“Emergency Handling for Recovery of SAP System Landscapes” and “Business Continuity
Management for SAP System Landscapes”, which you can find in the service marketplace
under https://service.sap.com/solutionmanagerbp
4.4.2 Troubleshooting
If executing this Best Practice did not produce the desired results
1. Search for related notes in the SAPNet.
2. Open an SAP Customer Message describing your problem

67
© 2008 SAP AG
68
Best Practice: Data Consistency Check
© Copyright 2008 SAP AG. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or for any purpose without the express
permission of SAP AG. The information contained herein may be changed without prior notice.
Some software products marketed by SAP AG and its distributors contain proprietary software components of
other software vendors.
Microsoft®, WINDOWS®, NT®, EXCEL®, Word®, PowerPoint® and SQL Server® are registered trademarks of
Microsoft Corporation.
IBM®, DB2®, OS/2®, DB2/6000®, Parallel Sysplex®, MVS/ESA®, RS/6000®, AIX®, S/390®, AS/400®, OS/390®, and
OS/400® are registered trademarks of IBM Corporation.
ORACLE® is a registered trademark of ORACLE Corporation.
TM
INFORMIX®-OnLine for SAP and Informix® Dynamic Server are registered trademarks of Informix Software
Incorporated.
UNIX®, X/Open®, OSF/1®, and Motif® are registered trademarks of the Open Group.
HTML, DHTML, XML, XHTML are trademarks or registered trademarks of W3C®, World Wide Web Consortium,
Massachusetts Institute of Technology.
JAVA® is a registered trademark of Sun Microsystems, Inc. JAVASCRIPT® is a registered trademark of Sun
Microsystems, Inc., used under license for technology invented and implemented by Netscape.
SAP, SAP Logo, R/2, RIVA, R/3, ABAP, SAP ArchiveLink, SAP Business Workflow, WebFlow, SAP EarlyWatch,
BAPI, SAPPHIRE, Management Cockpit, mySAP.com Logo and mySAP.com are trademarks or registered
trademarks of SAP AG in Germany and in several other countries all over the world. All other products mentioned
are trademarks or registered trademarks of their respective companies.
Disclaimer: SAP AG assumes no responsibility for errors or omissions in these materials. These materials are
provided “as is” without a warranty of any kind, either express or implied, including but not limited to, the implied
warranties of merchantability, fitness for a particular purpose, or non-infringement.
SAP shall not be liable for damages of any kind including without limitation direct, special, indirect, or
consequential damages that may result from the use of these materials. SAP does not warrant the accuracy or
completeness of the information, text, graphics, links or other items contained within these materials. SAP has no
control over the information that you may access through the use of hot links contained in these materials and
does not endorse your use of third party Web pages nor provide any warranty whatsoever relating to third party
Web pages.

68
© 2008 SAP AG

You might also like