Re: ENDSBS SBS(QINTER) OPTION(*IMMED) Takes 10 minutes?

On 29 Apr 2013 15:33, CRPence wrote:

On 29 Apr 2013 13:58, Sam_L wrote:

In day end processing on Saturday and Sunday we issue
ENDSBS SBS(QINTER) OPTION(*IMMED).
Then we DLYJOB(120)
Then we do STRSBS SBSD(QINTER)
This Saturday evening it took QINTER 10 minutes to end, so the
subsequent STRSBS SBSD(QINTER) failed.

<<SNIP>>

Examining recent QINTER job logs it normally ends in less than 12
seconds.

<<SNIP>>

Any idea why it took so long for QINTER to end?

<<SNIP>>
I did a little research with a web search of "PERFM" "ENDSBS"
<<SNIP>>

I performed a more liberal interpretation of the condition as a "performance" issue, using a little more variety in the likely keywords. Using "WAIT" [among others] I found the following:

http://www.ibm.com/support/docview.wss?uid=nas2334c7b41f21f3fd286257b220041f022
"Abstract

OSP-OTHER-WAIT INTERACTIVE SUBSYSTEM EXPERIENCES DELAYS WHILE SHUTTING DOWN

Problem Summary

QINTER experiences unexpected 10 minute delays randomly while shutting down

Problem Conclusion

The error occurs when one or more jobs end through an error path that does not do complete cleanup. This occurs before the ENDSBS is processed. When the ending of the subsystem reaches the point where secondary threads should end, some threads are still waiting for work that is implied by the incomplete cleanup. Cleanup is fixed to ensure that this error no longer occurs.

In addition, an error in handling interactive transferring jobs that come off of the job queue during the ending of the subsystem is fixed so that secondary threads are not left waiting.

...

PTFs Available

R710 SI49759 1000
http://www-01.ibm.com/support/docview.wss?uid=nas3a3785eb0f954258486257b42000089b7";