How do I determine a value for the average number of RIDS for calculating RID pool size?

Document ID : KB000021390
Last Modified Date : 14/02/2018
Show Technical Document Details

Introduction:

An application is causing RID Pool failures, so the RID pool size needs to change.  The formula for calculating RID Pool size is: Number of concurrent RID processing activities X average number of RIDs X 2 X 5 bytes per RID.

Question:

How do I determine a value for the average number of RIDS that can then be plugged into the IBM formula for calculating RID pool size?

Environment:

All IBM supported versions of DB2 for z/OS and z/OS

All GA releases of CA Sysview PMO for DB2 for z/OS

Answer:

 In order to determine a value for average number of RIDs, you need to run a trace to collect Ifcid 125, which is the MIAP-SELECTION record in CA Sysview for DB2. The trace should be qualified by fields that identify the application causing the problem, like Plan Name or Correlation Id.

One method would be to execute an Application Probe, option 8.2 from the CA Sysview for DB2 main menu, and set the "Include Detail" field on the Start Requests screen to 'Y'.  This will include the SQLTRACE IQL request in the Application Probe. The MIAP-SELECTION Ifcid record is collected as part of the SQLTRACE.

You can also create a custom IQL request that just collects the MIAP-COLLECTION (Ifcid 125) record. Use the section of the SQLTRACE IQL request that references this record to get started. Start the IQL request when the application is about to execute using the START "iqlname" command.

In both cases, you can complete various fields on the Start Requests screen to qualify the Probe or IQL request to collect data for the application causing the RID Pool failures. 

A third option is to start a DB2 trace to collect Ifcid 125. After dumping the systems' SMF datasets to a file, use that file as input to the Insight Batch Reporter. Use a custom IQL request for the MIAP-COLLECTION (Ifcid 125) record as the basis for the report.