WebSphere MQ-related problems, recovery, page set I/O error, data sets, example recovery scenarios, page set problem, CSQP004E message" /> Page set I/O errors

 

Page set I/O errors

Problem

A page set has an I/O error.

Symptoms

This message is issued:

CSQP004E +CSQ1 csect-name I/O ERROR STATUS ret-code
PSID psid RBA rba

System action

The queue manager terminates abnormally.

System programmer action

None.

Operator action

Repair the I/O error cause.

If none of the page sets are damaged, restart the queue manager. WebSphere MQ automatically restores the page set to a consistent state from the logs.

If one or more page sets are damaged:

  1. Rename the damaged page sets and define replacement page sets.

  2. Copy the most recent backup page sets to the replacement page sets.

  3. Restart the queue manager. WebSphere MQ automatically applies any updates that are necessary from the logs.

We cannot restart the queue manager if page set zero is not available. If one of the other page sets is not available, we can comment out the page set DD statement in the queue manager start-up JCL procedure. This lets you defer recovery of the defective page set, enabling other users to continue accessing WebSphere MQ.

When you add the page set back to the JCL procedure, system restart reads the log from the point where the page set was removed from the JCL to the end of the log. This could take a long time if a lot of data has been logged.

A reason code of MQRC_PAGESET_ERROR is returned to any application that tries to access a queue defined on a page set that is not available.

When you have restored the defective page set, restore its associated DD statement and restart the queue manager.

The operator actions described here are only possible if all log data sets are available. If your log data sets are lost or damaged, see Restarting if you have lost your log data sets.