AnsweredAssumed Answered

s722 abend in ESP master started task

Question asked by mbieganski on Apr 5, 2016
Latest reply on Apr 7, 2016 by don-t

Hello Community,   we're users of ESP v11.4 HAO one of our main groups have added a TON of recursive schedules and with the huge influx of more jobs running every minute find ourselves subject to S722 abends in the ESP master (too many lines).  We are going to start bouncing the master every 2 weeks (from its original 4 weeks) and that also will help alot.

The audit logs are much larger than in the past but we handle that by spinning off our audit logs to disk and they eventually are purged in sdsf so the audit logs are not the problem. 

What we see as the problem is the JESYSMSG in the ESP Master's started task, it can contain 6 or 7 million lines and to my eye not contain a single bit of info that we'd ever need to use,

(where as the jesmsglg and audit logs do contain valuable info which we consult almost daily).

 

In addition to the bouncing the master every 2 weeks to help keep the total lines down, we know we can write a jes exit and or add a jobparm L=nnnnnn to the masters started task to increase the default number of lines before the MVS system decides to S722 us, but thougth I'd ask out there if anyone faced this same issue and if there is some we to throttle that JESYSMSG that is running wild.   I think if we tried pointing the JESYSMSG to a msgclass that gets purged, the millions of lines would remain in sdsf while the started task is running so it would do us no good.

It just seems silly that primarily its the JESYSMSG lines that seem to contribute the most to the S722 abends and it provides little value.

Is there any way to turn jesysmsg off in the esp started task? Or point it to some msgclass that keeps it totally out of sdsf and thus its millions of lines will not count against us and cause S722 abends?

Outcomes