ITPAM Processes not being archived.

Products

• CA Process Automation Base

Issue/Introduction

We found that PAM Database is increasing - a result of processes not being archived. We tried changing the policy - in case it was a configuration issue. We also restarted the ITPAM service and the server, processes are still not being archived. We can successfully archive (and purge) processes manually. 

 

Environment

Release : 4.3

Component : Process Automation

 

Cause

This specific problem was related to a massive number of processes suddenly creating a bottleneck. This can be validated by using the queries below. Key conditions to validate whether this specific KB applies to your situation:

• SQL Select query: select count(*),CONTAINER from Node0ACTIVEMQ_MSGS group by CONTAINER
  Condition: EngineMgmtQueue had a large number of records
  Note: The number of DLQ records can usually be ignored. 
• SQL Select query: select count(*), runstate FROM dbo.c2oruntimeobjects where discriminatorType='WorkFlow' GROUP BY runstate;
  Condition: This query shows many records. Many records in this table is not normal. It is normal for this to be limited to a small number of records. The natural lifecycle of a process is for it to be moved from this table, when the process completes, to the c2oruntimeobjectsarch table. 

 

Important: 

It is critical that no manual changes are made to the database. These queries are informational queries only to help isolate problem/cause. Updating the DB directly can result in corrupting the database. 

 

Resolution

If you identify a bottleneck in the EngineMgmtQueue (EngineMgmtQueue only) then you can use the following steps to increase the number of threads (default 1) to help process the bottleneck in the EngineMgmtQueue:

• Add a "Common Properties" property named: noOfRollupConsumers; value: 5. This increases the number of threads used/responsible for processing the EngineMgmtQueue.
• Stop/Start ITPAM service.
• Evaluate if additional action is necessary. 
  Note: It may take time for the bottleneck to be resolved. Ideally, the number of records should not increase over a 24-48 hr period. If it is then please contact technical support.