An undetermined stall condition
Ask these questions when you are dealing with a stall condition.
- Questions to ask
-
- Did the stall occur during initialization?
- How far did initialization progress?
- Were there any definition or setup errors reported?
- Did the stall occur during operation?
- Are the necessary communication links between CMASs and MASs available?
- What type of request was being processed?
- How big was the CICSplex involved?
- How many CMASs and MASs were involved?
- What types of monitoring, real-time analysis, and workload management were active?
- Did the stall occur during termination?
- Did the stall occur in a CMAS?
- Did the request time out with an EYUEInnnn message?
The local CMAS may be waiting for one or more CICS® systems (or their CMAS) to return requested data. A CICSPlex SM view does not return until all the expected data is collected.
- Did the request time out with a CICS message?
- Did the request time out with an EYUEInnnn message?
- Did the stall occur in a MAS? Try stopping the MAS agent code (using the STOP action command from the MAS view), then evaluate the underlying CICS system.
- Is the CICS system taking an SDUMP?
- Is the CICS system looping or hung?
- Did the request time out with a CICS message?
- Is the CICS system experiencing a short on storage (SOS) condition, or has it reached its MAXTASK level?
- Did the stall occur during initialization?
- Documentation to collect
-
- System console log and EYULOG
- CMAS job logs
- Unformatted SDUMP from the affected address spaces (TSO, CMAS or MAS)