OPSLOG stopped getting messages after OPS3441O POSSIBLE PROCESS BLOCK DEPLETION, CODE= E (0 4128 0 )

Products

• OPS/MVS Event Management & Automation

Issue/Introduction

I had an issue over the weekend with OPSLOG missing messages for around 25 seconds (per timestamps in the OPSLOG) which caused OPS to miss important messages driving actions that resulted in availability issues.

The system was rather busy as we were starting many CICS regions.  

I saw a process block depletion message in the syslog, OPS3441O POSSIBLE PROCESS BLOCK DEPLETION, CODE= E (0 4128 0 ).

Here are the PROCESS related parms from 4.1.1 screen:
PROCESSBLOCKSIZE    1557K    
PROCESS             75 BLOCKS

Environment

OPS/MVS

Cause

The system had reached the maximum number of Process Blocks allowed, at which time logging temporarily ceases until the condition is received. 

Resolution

The number of Process Blocks can be increased via the OPS/MVS PARMS screen (=4.1.1).    Increasing the number of Process Blocks, should be done with caution, as storage constraints could potentially occur.  

In this particular issue the process blocks reached maximum due to a large number of initiators being started at the same time.  This condition has also been seen upon region shutdown when stopping a large number of initiators.   If increasing the number of available Process Blocks is a concern, a REXX program could be written that will issue the initiator commands in smaller batches, with a wait coded in between.   Below is an example of stopping initiators:   

   Call ISSUE_COMMAND cmdchar||"PI(1-20)"        
    t = OPSWAIT(2)
   Call ISSUE_COMMAND cmdchar||"PI(21-40)"        
    t = OPSWAIT(2)

......