CONCURRENT PROCESSING AND CONCURRENT MANAGER

Introduction
A user typically performs two types of activities in Oracle Applications—one online transaction and the other batch
processing. The batch processing job is also known as Concurrent Processing which allows the Oracle Application users
to schedule jobs in background while the user can work with online data entry operations.

When a user submits a report to be run as a concurrent request, the job enters in a request queue. The Concurrent
Managers continuously read request from this master queue and run the requests based on the request’s schedule,
priority, and compatibility rules.

Example:

Consider a simple example of extracting a report of all active users of applications from the database. You can submit
a standard seeded report named “Active users” provided within the application, and this report will generate the
output with the list of active users in the applications. The concurrent manager is responsible for taking input from the
users as a concurrent request, processing it on the required manager as a batch job, and generating results for the
submitted job

Terminology:
Concurrent request : Whenever there is a business requirement for generating a report or performing any business
transaction as a batch, then the Oracle end user will submit a request from the provided application responsibility. The
execution of the program for the report is the concurrent request.

Concurrent program : This is the program responsible for implementing the business functionality and generating the
required results.

Concurrent process : A concurrent program that is scheduled to run for performing a business transaction is known a
concurrent process, and the concurrent process will get associated with the OS process.

How the Service Manager Works As part of GSM, the service manager is responsible for starting and terminating
the concurrent manager and service processes. The service processes could be any process like web listeners, forms,
or any other application process that is part of GSM. If the internal concurrent manager (ICM) terminates on a node,
then the service manager running on that specific node will also terminate. The internal concurrent manager and the
service manager are linked to each other. As shown in Figure , the concurrent manager and other services are
distributed across three application nodes. Consider node2 is terminated because of some error; then the service
manager and internal concurrent manager running on node1 will migrate the process from the failed node to any
available active service manager node. ICM keeps on monitoring the processes running on all application nodes, and
if for some reason the process fails on any of the nodes, the ICM will try to restart this process. ICM will migrate the
processes from one node to another node only when there is a node failure.
Consider a situation where there a failure on node1; this node is one that is running the internal concurrent manager
process that is required by other nodes as well for proper functioning of the application services. ICM is monitored by
the internal monitor process (FNDIMON), and it keeps in constant communication with the internal concurrent
manager. If there is a failure because of a network connection or node failure, then ICM will be started on any one of
the surviving application nodes. You can configure the reviver process, which will monitor the connectivity. Whenever
there is a disconnection, ICM terminates itself and spawns the reviver process. Once the connectivity is established,
then the reviver starts the ICM and exits. You can start multiple ICMs on all participating application nodes, but all ICM
processes will get terminated, and only one will remain active—the one that was started first.

Figure shows the process ID of the service manager. PID 5590 is an operating system ID for the service manager that
is running on ERPNODE3.

Important Concurrent Managers

The key managers predefined by Oracle includes Internal Concurrent Manager (ICM), Conflict Resolution Manager,
and Standard Manager.

INTERNAL CONCURRENT MANAGER: Its main task is to ensure that all the other Concurrent Managers are up and
running. The ICM starts the FNDLIBR application process

[Note: The Internal Concurrent Manager is in turn monitored by Internal Monitor which is responsible for starting the
failed Internal Concurrent Manager in a local node. There should be an Internal Monitor defined on each node.]

CONFLICT RESOLUTION MANAGER: The Conflict Resolution Manager (CRM) takes care of resolving the program
incompatibilities and checks if a request in queue can be run in parallel with the running request.

STANDARD MANAGER: The Standard Manager is the master Concurrent Manager, This manager is always running and
it can take care of processing any concurrent request, it has no specialization rules. This manager runs 24 hours a day
for 365 days. It is responsible for running most of the business requests. This manager will communicate only with the
node service manager and client application processes. This concurrent manager will accept any requests that have no
specialization rules defined. The standard manager also uses the FNDLIBR application process
Transaction Manager: The transaction manager is responsible for taking the load from the concurrent request table
for pooling the requests submitted by the users, and these requests are sent to the standard manager. In a RAC
environment, it is required to activate the transaction managers on each cluster node.

Querying The Concurrent Manager

There are many ways to find out if the Concurrent Manager is running or not and a Apps DBA should be aware of all
the methods for checking if the Concurrent Manager is up and running.

1. Checking from the Operating System:

ps -ef | grep FNDLIBR |grep appmgr01
2. Checking from the forms:
The Concurrent Manager can also be checked from th e Forms by the following navigation > System
Administration > Concurrent > Manager > Administer.
If the actual is equal to the target it means the Concurrent Manager is up and running. The value of Actual
and Target should be greater or equal to 1. From the screen itself we can terminate, deactivate and restart
any of these managers.

3. Checking by running the sqlfile: The status of the Concurrent Manager can also be checked by running
afimchk.sql available at the FND_TOP/sql
4. Checking from Oracle Application Manager: We can also check for the Concurrent Manager by login into
the Oracle Application Manager and then by navigating to the Application Dashboard. The overview page
gives information about all the middle tiers components which are up and running. As shown in Figure there
is a mark in the concurrent processing tab. It means the Concurrent Manager is up and running.
 If we click the tick mark it will show in details which of the managers are up and running as shown in Fig

STOPPING CONCURRENT MANAGER

1. The stopping of Concurrent Manager is very simple

adcmctl.sh stop apps/apps

2. By querying for the process FNDLIBR and killing the same by issuing a kill –9
3. Individual Concurrent Managers can be deactivated from the Forms using the navigation > Concurrent >
Manager > Administer.
4. Using the CONCSUB utility, The syntax for the same is

$CONCSUB username/password SYSADMIN ‘System Administrator’ SYSADMIN WAIT=Y CONCURRENT FND ABORT
SUBMITTING CONCURRENT REQUEST
The concurrent requests can be submitted via forms using the navigation View > Request > Submit a New
request(System Administrator ➤ Concurrent ➤ Request). It then opens a window and then prompts for type of
request which you want to run a single request or a request set.

Multiple options are available on this screen, and each option is described here:

My Completed requests : Shows all requests completed by the user who is currently logged in

My Requests in Progress : Shows all requests that are still in progress

All My Requests : Shows all requests regardless of the phase for the user who is currently logged in

Specific Requests : Allows you to narrow down the search for concurrent requests with more specific information, as
shown in Figure

The option single request allows to submit only one single request at a time whereas the option Request Set allows to
run multiple concurrent programs and reports at one go. The request set can be further classified into two types—
private and public. Private request sets are created by the users who have not logged with system administrator
responsibility whereas public request sets are created by the system administrators. System administrators can also
update any request set, regardless of who the owner is.

Select the Single Request button and click ok. It opens the submit request form as shown in Fig , Select the program
which needs to be run from the list.

lists the standard requests that are available within that responsibility
You can select any request from the available list,

The request can either be submitted immediately or can be scheduled for a later period of time. Click the button
schedule for scheduling the same for a later period of time. By default the concurrent request are submitted
immediately.

The concurrent program can be run immediately, only once, periodically or on some specific days. You can also save
this schedule for future reference and can use the same schedule for a different Concurrent Program by using the
option apply a Saved Schedule.

The Completion option refers to what Oracle Application will do once the request is completed. It can notify people
via email, can save the output in a file, can take a print out of the same or simply won’t do anything.
There are multiple options available for the concurrent request, as shown in Figure . The following list explains the
options:

Hold Request : Any request that is in pending state can be kept on hold. The system will not run these requests until
the hold status has been removed.

View Details : This will provide detailed information about a specific request.

View output : This will provide the output file for the executed concurrent request in HTML, text, PDF, and so on.
Cancel request : Any scheduled or running request can be cancelled.

Diagnostic : This will show the start time and end time of the particular concurrent request.

View Log : This will show the log file for that particular concurrent request, which can be helpful in troubleshooting
the concurrent requests that complete with a status warning or error.

Request Arguments
Oracle concurrent requests have multiple options available to support the business needs. These options, listed in
Figure , are generic for all concurrent requests
Copy : This will allow you to copy any previously completed concurrent request

Language Setting : This will provide an option to choose the languages in which the report must be generated.

Schedule : This is an important feature of the concurrent manager. The concurrent requests can be scheduled to
execute at some specific time, and they can be scheduled to run at periodic intervals

By default the value is As Soon as Possible . This means the request will run immediately if the manager is available,
or it will wait for the next manager to be available.

Once. This means you can schedule the execution of the concurrent request for a specific date and time
The concurrent request can be scheduled to run periodically , You can provide the start and end dates for the specific
schedule, and you can provide the periodic parameter Re-run every x Month(s), Week(s), Day(s), Hour(s), Minute(s).
The option “Increment date parameters each run” will increment to the subsequent value based on your selection of
the “Re-run every” parameter.

The On Specific Days scheduling option is simple; you can choose any specific calendar day or any day of the week for
the provided period

Once the request is submitted it gives a request ID and prompts if we want to submit a new request.

THE CONCSUB UTILITY:

The concurrent request can also be submitted from the command line syntax using the Concurrent Submission utility
CONCSUB.

Syntax: CONCSUB <ORACLE ID> <Responsibility Application Short Name>

<Responsibility Name> <User Name> [WAIT=<Wait Flag] CONCURRENT <Concurrent

Program Application Short Name> <Concurrent Program Name> [START=<Requested

Start Date>] [REPEAT_DAYS=<Repeat Interval>] [REPEAT_END=<Request

Resubmission End Date>] <Concurrent Program Arguments ...>

Example: CONCSUB SCOTT/TIGER SYSADMIN ‘System Administrator’ SYSADMIN WAIT=Y

CONCURRENT FND FNDMNRMT START=’”01-JAN-2000 23:00:00"’ REPEAT_DAYS=1

REPEAT_END=’”01-JAN-2001 23:59:00"’ Y 0 0
Standard Requests, Custom Requests, and Request Arguments
Oracle delivers a set of seeded concurrent programs for generating reports and performing business transactions,
and the requests that use these standard programs are known as standard requests.

There may be certain business requirements that a standard manager is not capable of performing. In that situation,
you may need to create the custom concurrent program, and the request that will use this custom program for
performing a business transaction is known as the custom request .

Concurrent requests need a certain set of arguments for processing the business transaction, and these arguments
are purely specific to the concurrent requests and will not be the same for all requests. The arguments that are
passed on by a user to a concurrent request are known as request arguments

FLOW OF A CONCURRENT REQUEST:

1. Once a Concurrent request is submitted by the user, the table FND_CONCURRENT_REQUESTS is automatically
updated with the details of the request.
2. The table is also updated with the information about the schedule of the concurrent request weather its
immediately scheduled or scheduled at a fixed time.
3. Once the request is scheduled to run the Concurrent Manager checks the FND_CONCURRENT_TABLES to find
out if the request is incompatible with any other request.
4. If the request is incompatible then the Conflict Resolution Manager takes care of the request and find out what
are the incompatibilities and it resolves the incompatibilities.
5. If there are no incompatibilities then it’s checked weather any special manager is there to take care of this
request. If there is any special manager to take care of this request then it goes to the queue of that manager
else the standard manager takes care of the same. Once the request is processed the
FND_CONCURRENT_REQUEST table is updated with the status.

COMBINATION OF PHASE AND STATUS

The Completed Phase:
This phase has five status which are discussed below

• Normal: The request completed normally.

• Error: The request failed. The details about the error can be checked from the logfile.
• Warning: The request completed with a warning. The warnings are generated when the upon completion task
don’t complete successfully.
• Cancelled: The request was cancelled before it was started.
• Terminated: When a running request was terminated.

The Running Phase

This phase has four status which are discussed below

• Normal: This means that the request is running normally and no interfere is required.
• Paused: This means that this request is waiting for some other request to be finished. This usually happens
when the completion of this request depends on the completion of the other.
• Resuming: This means that the paused request is resuming. This takes place once the dependent request
completes and the request which was paused can now be taken care.
• Terminating: This happens when the user or sysadmin chooses to terminate the request from the forms by
clicking the terminate button.

The Pending Phase

This phase has four status which are discussed below

• Normal: This phase is very common which means the requests are waiting for the Concurrent Manager.
• Standby: If the concurrent request is incompatible with some other request then they are put in the standby
status till the Conflict Resolution Manager resolves the incompatibility and the manager is allotted.
• Waiting: This usually happens when this request depends on completion of some other requests so that once
the other request is finished the output for the same can be used to complete this request.
 • Scheduled: This means that the concurrent request are scheduled to run at a point of time in near future.

The Inactive Phase

This phase has three status which are discussed below

• : This means either the Concurrent Manager is down or no specialized manager is defined to run this
request.
• Disabled: This means that though the concurrent program has been requested but it has not been enabled
yet.
• On Hold: This means that a pending request has been kept on hold for the time being.

DEFINING CONCURRENT MANAGERS

For defining Concurrent Managers we need to login to the ebusiness suite as sysadmin and need to choose the System
Administration responsibility. Then select the tab Define under the heading Concurrent Manager in the right hand side.

The exact screen shot of the same is shown in Fig

Adding new Concurrent Managers is one of the crucial decisions which the Apps DBA needs to take. Before adding new
Concurrent Managers a lot of factors need to be examined. One of the major area which needs to be analyzed properly
is the resource.

If you have less number of managers then your over all concurrent processing gets delayed with a lot many requests
having the status pending. On the other hand if you have too many managers then there will be a heavy load on your
operating system and online transactions will suffer which would result in performance issues. Thus before adding any
new managers a proper analysis needs to be done to obtain the optimum performance.

Once clicked on the Define Concurrent Manager link the forms launch and the window for defining the Concurrent
Manager opens.
Manager: The first field is Manager in which the name of the Concurrent Manager needs to be given which you
would be adding.

Short Name: The second field is Short Name in which a short name of the Concurrent Manager needs to be given

Application: The Application column helps to identify which application this Concurrent Manager would be running
but this doesn’t mean that this Concurrent Manager won’t be able to run programs associated with other
applications

Description: In this field put short description of this Concurrent Manager

Type: In this field you need to define what type of Concurrent Manager you are going to define. You cannot update
this field once defined. There are several types which are available at the LOV which are Concurrent Manager,
Transactional Manager, Internal Monitor etc. Select the manager which you need to define

Cache Size: This refers to the number of requests the manager remembers each time it reads which requests it needs
to run. If a manager has a cache value of five then it will read five requests at a time and wait till these five requests
have been completed before it starts any new request. This parameter is used to tune the Concurrent Manager to
work in an effective manner

Note: Oracle suggests to enter a value of 1 when defining a manager that runs long and 3 or 4 for the managers
which run small quick jobs

Data Group: This is used only by the Transaction Manager. This is used by the transaction manager to connect the
database

Consumer group: Resource consumer groups and resource plans provide a method for specifying how to partition
processing resources among different users

Parallel concurrent processing details:

• Node: This field is applicable only if you have enabled parallel concurrent processing. Here you define the
node from which your manager will operate. The node must be registered with the Oracle Applications and
an entry of the node should be there in the table FND_NODES
• System Queue: If you are operating in a parallel concurrent processing environment and you want your
manager to use a platform-specific queue management system instead of generic concurrent processing
queue management, specify the queue or class name of that system.

Program Library
Name: You need to assign a predefined Library for your manager. The Concurrent Manager runs only those programs
which are listed in their program libraries. If the specialization rule includes any other type of concurrent programs
then the Concurrent Manager can also run those programs. The following are a few libraries.
Specialization Rules
From here you can specialize your manager to run only certain types of requests. If you don’t define specialization
rules then the manager can process any kind of concurrent request.

• Include: Use this drop down button to include or exclude those requests which your manager will/won’t run.
• Type: Here you specify the type of specialization rule you want to assign to your manager. There are five
types which are available in the drop down list viz Combined Rule, Oracle ID, Program, Request Type and
User.
• Application: The Application refers to the various Applications which will be either included or excluded for
processing by this Concurrent Manager. You can add as many applications you want this manager to take
care.
• Name: This is the actual name of the concurrent program which will be either included or excluded by this
Concurrent Manager.

The Specialization rule screen is shown in Fig

Work Shifts
The Work Shift defines the time for which the Concurrent Manager is active. You can define some fixed date or time
for or can make the manager run 24*7 making it active all the times. The Work Shifts are defined by using the Work
Shift form
The Process tab defined the number of the operating system process which will run to process the concurrent
requests. Each process can run a concurrent request.

The Parameter tab is used for Generic Service Management.

The Sleep Seconds is the sleep time for your manager during the work shifts. This is the number of seconds the
Concurrent Manager waits between checking the list of pending concurrent requests.

DEFINING WORK SHIFTS

For defining the workshift for the first time follow the following Navigation > Concurrent > Manager > Work Shifts

Once the Work Shifts button is pressed it opens a new window from where we can define the various work shifts
which we can use while defining new Concurrent Managers

We can define as many work shifts we want and can define a work shift to run at any particular day or time.

• Name: This refers to the name of the work shift which you will be defining.
 • From/To: This is the time at which your concurrent shift begins and ends.
• Days From/To: This refers to weekdays from which your concurrent shift will begin and end. From the figure
given above we can see the shift of ABC starts from Monday and ends in Friday.
• Date: Enter a date if you want to define a date specific work shift

ADMINISTRATING CONCURRENT MANAGERS

You can administer any Concurrent Manager from the screen below

1. Name: This is the first column which contains the name of the Concurrent Manager. We can see from the
figure the highlighted one is standard manager.
2. Node: Node refers to the node from which the Concurrent Manager is started. In parallel concurrent
processing where you configure multiple node the node information is obtained from this column.
3. Process Actual: The process actual refers to the actual number of the processes running for that Concurrent
Manager. Each process can run one concurrent request. So normally for Standard manager the process is
always greater than all the other managers.
4. Process Target: The process target refers to the maximum number of the processes that can be active for
this manager. Sometimes the actual process may be less than of Target process due to manager deactivation.
5. Request Running: This shows the number of the requests currently being processed by the Concurrent
Managers.
6. Request Pending: This shows the number of requests which are pending are waiting for the Concurrent
Managers to complete the running requests. If the actual process for the Concurrent Manager is less and too
many concurrent requests are submitted together then this number goes up.
7. Status: This field shows the status of the manager after you have chosen a particular action which are shown
in the buttons below. For example if you select a particular manager and clickon the restart button below
then the status field will have the value restarting.

CONTROLLING SPECIFIC MANAGERS

1. Terminate: When you terminate Internal Manager then all the managers are automatically gets deactivated
and all the running requests are terminated. If you want to terminate a particular manager then select the
manager and click the button terminate. The status of the manager changes to deactivate after a few seconds
and all the requests processed by that manager are immediately terminated. Once a manager is terminated it
doesn’t restart automatically, you have to manually restart the same using the Restart button.
2. Deactivate: For deactivating a particular manager select the manager and press the button deactivate. In case
of deactivation all the requests processed by the manager are allowed to complete before the manager shut
down. If you deactivate the Internal Manager then all the managers automatically gets deactivate but all the
running requests are allowed to complete before the manager is shut down. This is the only difference
between termination and deactivation. In termination all the running requests are terminated immediately
whereas in case of deactivation all the running requests are allowed to complete at first.
3. Restart: The restart option is available when you select a particular manager. With the restart button you can
restart a particular manager. Whenever you change the definition of the Concurrent Manages you need to
 restart the same in order to make the change effective. You may also need to restart if you have changed the
work shift for a particular manager or you modify the number of the target processes. In case of parallel
concurrent processing if you change the Node of a particular manager then also you have to restart the
manager in order to reflect the change.
4. Verify: The verify button becomes enable only when you select the Internal Manager which means that this
option is only available for the Internal Manager. One of the functions of the internal manager is to monitor
the processes of each Concurrent Manager. The process of monitoring the other Concurrent Manager by
internal manager is known as the PMON cycle. When you click the verify button you can force the process
monitoring or the PMON activity to occur.
5. Suspend: The suspend option is available only for services managed by Generic Service Management (GSM).
The suspend option suspends the operations of the service.
6. Resume: Like Suspend this option is also available for services managed by GSM. The resume option resumes
the operation of the service.
7. Refresh: This button refreshes the current window and once this button is clicked you can get the updated
status of the manager. Suppose you have selected a manager and have clicked the control button terminate
then by pressing the refresh button you can come to know when the manager gets actually terminated.
8. Request: This button will open the Request Submission Form which we had already discussed earlier with
which Concurrent Requests can be submitted.
9. Process: The process button shows the details of the processes of the given Concurrent Manager. It displays
all the processes which are active, terminating, migrating as well as those processes that have been terminated
or deactivated.

IMPORTANT DIAGNOSTIC SCRIPTS

The following SQL scripts located under $FND_TOP/sql are useful when diagnosing Concurrent Manager

problems:

1. afimchk.sql: Tells the status of the ICM and PMON method.

2. afcmstat.sql: Lists active manager processes.

3. afrqrun.sql: Lists all the running, waiting and terminating requests.

4. afrqwait.sql: Lists requests that are constrained and waiting for the ICM to release them.

5. afrqscm.sql: Prints log file name of managers that can run a given request. It can be used to check for possible errors
when a request stays in pending status. It requires a request ID value.

6. afcmcreq.sql: Prints the log file name of the manager that processed the request.

7. afrqstat.sql: Summary of completed concurrent requests grouped by completion status and execution type. It
requires number of days prior to today on which to report parameter.

8. afimlock.sql: Lists locks that the ICM is waiting to get.

9. afcmrrq.sql: Lists managers that currently are running a request

Working with Concurrent Request Sets
you will see how to create a request set using the Request Set Wizard(System Administrator ➤ Concurrent ➤
Requests ➤ Set)

Click on Request Set Wizard, two options are available to choose from: whether the requests in a request set should
execute in a sequential fashion or in a parallel fashion. Individual requests will be executed based on this selection. In
sequential request sets, the next request will start only after completion of the prior request. In parallel, all requests
will be started in parallel.

Figure shows two available options that a request set can perform in case a concurrent request completes with the
status Error. Say that there are three requests in a request set and each is request is dependent on other request
successfully completing. In such cases you should select the option Abort Processing. If there are no dependencies
between the requests, then you can choose Continue Processing. So, based on the nature of requests and
dependencies between the concurrent requests in a request set, you should select an appropriate option.

Above Figure requests the input for creating the request set.

Set : Any desired name can be provided, but it’s always recommended that you use a relevant name.
Application : You should select the registered application.
Description : This is a brief description about the requests and request set.
Note For custom applications it’s always recommended to define requests/request sets with names that start with
XX_Custom_request_set.

Next figure shows how to configure the printing option for the request set. You can print requests for each individual
request in a request set, or you can print all requests together after the successful completion of a request set.

Next figure shows how to add individual requests in a request set. Here we have selected two concurrent programs:
Active Responsibilities and Users and Active Users

Figure 1 include requests

Next figure confirms the creation of the request set.

Next figure displays the created request set and also displays other actions that can be configured for a request set
such as Define Stages and Link Stages.

Next Figure shows the sequence of execution of the individual requests. The request with sequence 10 will start first,
and the processing will continue based on the incremental sequence numbers.
Next figure shows the Link Stages option of a concurrent request set.

Let’s see the execution of a request set. The submission process of a request set is similar to a single request. Figure
shows the selection of the Request Set option.

Next figure show the submission screen of newly created request set.
Once you click Submit, it will generate the new request ID for the submitted request set. This ID will be identified as
the parent ID, and all requests running under that request set will be identified as child requests.

Managing Concurrent Manager Incompatibilities:

Concurrent program incompatibilities are managed by conflict domains. The following are the possibilities of
concurrent manager incompatibilities:

• A concurrent program can defined to be incompatible with a specific concurrent program or with multiple concurrent
programs.

• A concurrent program can be defined to be incompatible with itself so that it will not allow two of the same programs
to run simultaneously.

• A concurrent program can be defined to be incompatible with all other programs, so that program will run alone.

• A concurrent manager incompatibility can be defined for request sets.

We’ll show you a simple demonstration for defining a concurrent manager incompatibility. We will use the same
request set we created earlier in this chapter. To navigate to the right spot, select System Administrator ➤ Concurrent
➤ Set and search for the request set called Apress Request SET.

In next figure you can see the option Allow Incompatibility. If you check that box, then the incompatibility will be
defined for that concurrent request set. After saving the record, the system will generate a new request set with the
name Request SET Apress Request SET. This newly created request set will allow you to choose other concurrent
programs for defining incompatible rules.

To navigate here, select System Administrator ➤ Concurrent ➤ Program ➤ Define and search for Request Set
Apress Request SET. Then click Incompatibilities.

Next figure displays the newly created concurrent request set on which you can define the incompatible concurrent
programs.
You can define as many incompatible concurrent programs as possible for Request Set Apress Request SET, as shown
in Figure

The concurrent program can be a request set or individual program, and it can be selected from the Scope option, as
shown in the above figure

Two options are available for managing the incompatibility for the concurrent program. If the Domain type of
incompatibility is selected, then it will be resolved at the domain level. If Global is selected, then it cannot run with
any other program globally irrespective of domain. The Domain option will identify the list incompatibilities that a
concurrent program has to resolve between programs.

■ Note All concurrent programs use the standard conflict domain. If the custom profile option Concurrent: Conflict
Domain is set, then it will be defined using the program parameter.

Creation of Custom Concurrent Manager

The concurrent managers installed by default are known as standard concurrent managers . Oracle EBS allows you to
create custom concurrent managers to meet your business requirements. Consider a situation where you want your
request (Request A) to be processed within five minutes and your standard managers are busy serving other requests.
Request A, which has priority, eventually needs to wait until other requests are completed and the next standard
manager process is available to serve Request A. If you create a custom concurrent program and associate Request A
with the custom program, then the request will complete immediately because the manager is available for Request
A only. In this section, you will see how to create a custom concurrent manager. To create custom manager, navigate
to System Administrator ➤ Concurrent ➤ Manager ➤ Define. On this screen you have to provide the required
parameters for creating the custom concurrent Manager.

On this screen you have to provide the required parameters for creating the custom concurrent Manager

Manager : This can be any desired name, but you have to ensure you use XX with the manager name so that it is
identified as a custom manager.

Short Name : Any relevant name should be provided without spaces.

Application Name : You can select any application name for which the custom manager is being created.

Type : This will list all the possible types of services of the GSM, and you can select any value based on the requirement.

Program Library : Here you can use any other custom library as well if created.

Below figure shows all the required parameters for creating a custom concurrent manager. After inputting all these
parameters, you should save the records.

Once the creation of the concurrent manager completes, then you have to define specialization rules and work shifts
for this custom concurrent manager. In this example, we have provided the standard work shift, which will run 24
hours every day with three processes. Figure below shows the work shift of the custom manager XX Apress Manager

Now you will define the specialization rule for the newly created concurrent program. You will define a rule for this
manager to include only the Active Users concurrent program. Figure below shows the specialization rule for XX Apress
Manager

In the next section of this chapter, you will see details of how you can create the custom work shifts and specialization
rules for effective utilization of resources. You can see the newly defined custom concurrent manager using this
navigation: System Administrator ➤ Concurrent ➤ Manager ➤ Administer.
By default the manager will be deactivated, so you need to activate it from the same Administer Concurrent Manager
screen.

Working with the Specialization Rule and Work Shifts

In concurrent processing, each concurrent manager is associated with a work shift and specialization rule. By default
Oracle will have only a standard work shift that will run 24 hours every day. If you do not specify any specialization
rule, then it will accept all concurrent requests.

The purpose of using work shifts and a specialization rule is to define a set of concurrent requests that should run on
specific managers on predefined timings. Here you will see how to create the custom work shifts and custom combined
specialized rule and then assign these to the required concurrent manager. To create a custom work shift, navigate to
System Administrator ➤ Concurrent ➤ Manager ➤ Workshifts.

In the Name section, you should provide the work shift name and provide the time for which the work shift should be
active. Figure 10-42 shows two newly added work shifts: Night Data Loading and Weekend. These work shifts will be
activated only during the specified days and time as defined.
Combined specialization rules will combine one or more rule to generate a single rule.

To create a combined custom specialized rule, navigate to the following: System Administrator ➤

Concurrent ➤ Manager ➤ Rule.

Combined Rule : Provide any desired name.

Application : Provide any registered application for which the rule is being created.

Description : Provide relevant information for the rule.

Specialization Rules : Here you can include and exclude programs, users, Oracle IDs, request types, and you can add
multiple records based on the business requirement.

Figure below shows the specialization rule that includes two concurrent programs and excludes one user, so this rule
will accept only these two listed concurrent programs and excludes user FINADMIN for running these requests.

Now you will add the newly created work shift and combined specialization rule to the custom manager.

Below Figures show how to assign the custom specialization rule and work shift to the required concurrent manager.
Modifying specialization rules requires the concurrent manager to be restarted. So, you must consider whether it’s
possible to reboot the managers after modifying the specialization rules.
The XX Apress Weekend Manager will activate only on the weekend, as defined in the Work Shifts window, and it will
run only one process with this work shift

Troubleshooting Concurrent Manager Issues

Proactive monitoring is always recommended for avoiding issues.

The first step is to ensure all actual and target processes are the same for the defined concurrent managers.

Figure below is the concurrent manager’s administer form. The Target column shows the defined value of the
concurrent manager, and the Actual column shows the actual running processes for that concurrent manager. If
there is a difference in the process count between the two columns, then you must investigate the relevant manager
log files
To view the log files, navigate to the following location: System Administrator ➤ Concurrent ➤ Manager ➤
Administer. Select the respective manager and click Processes.

From the Back End

You should check the $APPLCSF/log directory for a specific manager process log file, Here check the log file for
concurrent process 8281

Concurrent Manager Process log File

Internal Manager Log File

You should search for log file starting with the instance name in the same directory as $APPLCFS/log
This is the internal manager log file, and it records all the information for other concurrent managers.

Using Concurrent Manager Recover Wizard

The following notes are helpful for using the Concurrent Manager Recovery Wizard.

• “Concurrent Processing - Command-Line Utility OR Recovery Wizard” (Doc ID 134007.1)

• “Concurrent Manager Recovery Wizard - Oracle Applications Manager Troubleshooting and Diagnostics” (Doc ID
2130545.1)

This is how you navigate to the Concurrent Manager Recovery Wizard: Oracle Application Manager ➤

Site Map ➤ Diagnostic and Repair ➤ Concurrent Manager Recovery.

Figure below shows the navigation path for using the Concurrent Manager Recovery Wizard.

This is the first screen of the Concurrent Manager Recovery Wizard. To start the wizard, click Run Wizard, as shown
in Figure

Problems with Printing Reports

The Oracle Support tech note “EBS Reports & Printing Analyzer” (Doc ID 1610143.1) will be helpful for
troubleshooting report-printing issues.

Proactive Maintenance
To ensure that the concurrent managers work 24/7, you need to monitor and perform proactive

maintenance on Oracle EBS systems.

1. Ensure enough free space is available for the concurrent manager log/output files and also that there is enough
space available in the FND_FILE (utl_file package) directory.

2. Ensure all FND diagnostic profile options are disabled at levels.

 3. Schedule standard concurrent purge requests with the correct age parameter as per the SLA and business
requirements.

To schedule and the purging concurrent request, please refer to these MOS tech notes:

• Concurrent Processing - How to Optimize the Process of Running Purge Concurrent Request and/or Manager Data
(FNDCPPUR) (Doc ID 92333.1)

• Concurrent Processing - Purge Concurrent Request and/or Manager Data Program (FNDCPPUR) (Doc ID 104282.1)

4. Ensure there is enough free space available on the database tablespaces, especially APPS_TS_TX_DATA and
APPS_TS_TX_IDX.

5. Monitor pending concurrent requests on the respective managers. If the value is increasing for pending requests,
then increase the processes for the respective manager. But before increasing the concurrent manager processes,
you should evaluate the existing hardware resources. For example, if you have a concurrent node with 12 CPUs, then
you can define 10 processes for the concurrent manager. It’s always recommended to maintain room for other
operations.

These are some of the key areas that you should focus on for troubleshooting and proactively managing the
concurrent managers. The following MOS tech notes are specific for release 11i and 12.1, but the information listed
in these document is still valid for troubleshooting issues on 12.2.

• “Concurrent Processing - Creating Concurrent Manager Diagnostic and Debug files (troubleshooting)” (Doc ID
1312980.1)

• “Concurrent Processing - Troubleshooting Concurrent Manager Issues (Unix specific)” (Doc ID 104452.1)

Important Concurrent Manager Tables

Managing FMW/WebLogic
Introduction to Fusion Middleware in EBS 12.2
Oracle Fusion Middleware is a wide collection of software providing a range of services, such as Java EE, developer
tools, integration, identity management, business intelligence (BI), and collaboration. In addition to these services,
FMW supplies core services for the software stack placed on it, such as concurrency, transactions, threading,
messaging, and the SCA framework.

With this software infrastructure and technology , FMW can be considered an application server, a BI platform, a
content management repository, an SOA, a service bus called Integration and Business Process Management (BPM ),
a security and identity management platform, and a web portal. In addition, FMW offers a sophisticated
management platform for managing the components provided with it.

EBS 12.2 uses the web server and application server components of FMW. The rest of the services EBS provides rely
on other technologies such as Oracle Forms, Oracle Reports, and EBS product-specific software codes.

FMW components , Oracle HTTP Server, and WebLogic Server are standard FMW components in EBS 12.2. They are
not specialized or customized for use with EBS. The only difference between the FMW components used in EBS 12.2
and the stand-alone FMW components is that EBS 12.2 has AutoConfig on top of these FMW components.

With Release 12.2, Oracle has replaced the 10.1.3 Oracle Home that was used by EBS 12.1 with Oracle HTTP Server
(OHS) and WebLogic Server (WLS), which are parts of Fusion Middleware 11g.

Architecture and Components