Problems with long-running units of work

Use this topic to investigate, and resolve problems with long-running units of work.

This topic explains what to do if you encounter a long-running unit of work during restart. In this context, this means a unit of work that has been active for a long time (possibly days or even weeks) so that the origin RBA of the unit of work is outside the scope of the current active logs. This means that restart could take a long time, because all the log records relating to the unit of work have to be read, which might involve reading archive logs.

Old unit of work found during restart

Problem

A unit of work with an origin RBA that predates the oldest active log has been detected during restart.

Symptoms

IBM MQ issues the following message:

CSQR020I +CSQ1 OLD UOW FOUND

System action

Information about the unit of work is displayed, and message CSQR021D is issued, requesting a response from the operator.

System programmer action

None.

Operator action

Decide whether to commit the unit of work or not. If you choose not to commit the unit of work, it is handled by normal restart recovery processing. Because the unit of work is old, this is likely to involve using the archive log, and so takes longer to complete.