Mark, I honestly cannot respond as to how to investigate CATALOG-specific problems. I hope someone else can respond to that part of your question.
Looking at the numbers you have from the PFA check and Runtime Diagnostics results below, it looks like CATALOG is really very consistent across all 3 time periods for the rate of enqueues per the amount of CPU used. So, PFA determines that the really low "current rate" is unexpected and significantly lower than normal. Like you said, PFA then calls RTD and if RTD corroborates that there might be something wrong with that address space, the exception is issued. In order for RTD to issue the enqueue contention event, the address space would have had to have been waiting for > 5 seconds. A few questions for you: 1) How often does this occur? 2) Does it occur randomly or at a certain time of day/week/month? 3) Does it occur when a certain workload is running? 4) Do you have all the PFA and RTD PTFs for R13? (I don't know of any particular fix that would change your results especially since the results look valid, but I always encourage everyone to keep up-to-date with PFA PTFs.) I cannot answer the question as to why CATALOG would be in enqueue contention to determine if this is a real problem or "just the way CATALOG works" in your environment. After you investigate it, if you determine that "it's just the way it works" and you'd prefer to avoid the exceptions by having PFA ignore that address space, you can put CATALOG in the EXCLUDED_JOBS file for that check. Karla Arndt IBM PFA and Runtime Diagnostics for z/OS ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
