IMS is not operational

Problem

IMS is not operational.

Symptoms

More than one symptom is possible:

• IMS waits or loops

  WebSphere MQ cannot detect a wait or loop in IMS, so find the origin of the wait or loop. This can be IMS, IMS applications, or the IMS adapter.

• IMS terminates abnormally.

  • See the manuals IMS/ESA Messages and Codes and IMS/ESA Failure Analysis Structure Tables for more information.

  • If threads are connected to WebSphere MQ when IMS terminates, WebSphere MQ issues message CSQ3201E. This message indicates that WebSphere MQ end-of-task (EOT) routines have been run to clean up and disconnect any connected threads.

System action

WebSphere MQ detects the IMS error and:

• Backs out in-flight work.

• Saves in-doubt units of recovery to be resolved when IMS is reconnected.

System programmer action

None.

Operator action

Resolve and correct the problem that caused IMS to terminate abnormally, then carry out an emergency restart of IMS. The emergency restart:

• Backs out in-flight transactions that changed IMS resources.

• Remembers the transactions with access to WebSphere MQ that might be in doubt.

You might need to restart the connection to WebSphere MQ with the IMS command:

/START SUBSYS subsysname

During startup, IMS requests the adapter to resolve in-doubt units of recovery.