Parameter settings to avoid OSF looping in OPS/MVS

Products

• OPS/MVS Event Management & Automation

Issue/Introduction

I ran a REXX and it looped on one of the OSF servers, I remember in a previous installation I had found an article on what parms to update to keep this from happening.    I think two of the parms are OSFCPU and COMMANDRATE.  Please let me know what parms should be updated are and what the settings should be.   

 

 

Cause

OSF Looping is often caused directly by the parameter settings being less than optimal for the environment in which it is running.  

Resolution

The first step should be to ensure that the latest OSF maintenance is applied.  

It order to optimize the parameter settings, certain information about the environment has to be known. It is difficult, at best, for Broadcom to provide values for OPS/MVS tuning/threshold types of parameters when we are not at all familiar with a shop's automation code and data.  But  here are some overall factors to consider: 

• When executing REXX exec's, whether in OSF servers or other address spaces, the REXXMAXxxxxx parms control upper limits and they can be overridden for individual execs with the OPTIONS instruction. Most of these parms control how many of something, such as commands or SAYs, can be done from one exec. 
• The OSFxxx parameters mostly control execution type items like CPU time and wall clock run time.  These parms can not be overridden for individual execs and is the reason why we provide different types of OSF address spaces.  
• Also, the more data that an exec works with, like RDF tables and EXECIO data, the more time and resources the exec will need.- this is the silent one that creeps up because more and more data is kept and referred to.  Consider that the more CALLs an exec makes, the larger the code path becomes leading to an increased possibility of exceeding the REXX parms. 

Additional Information

For documentation on the OSF parameters see the "OSF Parameters" section of the Reference Guide.