Message is occurring four times daily consistently about the same time of day. OPS/MVS is active on the LPAR but is not actively managing work yet. All 300+ resources are in PASSIVE mode and the 20+ time-of-day schedules are not enabled. We are about to start actively managing all resources and 20+ time-of-day schedules.
Is this a problem?
Release : 13.5
Component : OPS/MVS
If using OPS 13.5, the PTF SO13481 may be the fix if the PROCESSUSE parameter shows values around 2147483643. Visit the OPSVIEW panel 4.1.1 to check the value of this parameter.
* For the situation where the message OPS3441O POSSIBLE PROCESS BLOCK DEPLETION, CODE= E (0 2356 0 ) occurs on a repeated cadence each day:
The OPS3441O and HealthChecker warning messages are both associated to this running out of process blocks condition - although we only observed the high evaluations condition for the OPS3441O message. We recommend to 'stagger' any automation that can cause this - ie. drain 30 initiators at a time in place of ALL the initiators, consolidate MSG* rule processing, etc. Some customers may instead opt to just increase the number of PROCESS blocks for the immidiate relief; however, it is strongly encouraged that the processes that drove the Process Block depletion still be investigated.
* For repeated OPS3441O POSSIBLE PROCESS BLOCK DEPLETION, CODE=B and CODE=E messages occur but the PROCESSUSE, PROCESSHIGHCOUNT do not look high:
The ProcessHighCount indicates the high amount usage, but pay attention to the value of what the ProcessASIDMax is. In this instance it was set to 15, so any one asid that uses 15 processblocks is subject to losing automation. Check if the application/s that issues the messages reached this value (asidmax). The OPS3441O messages are indications that possible process block shortage conditions may be occurring, further investigation should be done. In one case, investigation revealed that at least 4 process blocks were to be used for each WTO issued, one for each MSG * rule, and the customer had multiple MSG* rules issued at a time. The recommendation was to review the need for MSG * rules and consolidate ONLY needed code into one MSG * rule, that is coded efficiently with early out conditions and SELECT/WHEN/OTHERWISE. Minimal temporary relief can be achieved by increasing Process blocks.