> SRBs are interesting creatures. I had one recently that would takeanywhere > from 50 microseconds to 50 MILLISECONDS to be SCHEDULEd inside disabled
> code. I have a hard time understanding why it takes so long to put > an SRB on a > queue. SCHEDULE and IEAMSCHD generally do not put the SRB on a queue They obtain a WEB and put the WEB on a dispatching queue in priority order. Obtaining a WEB could occasionally result in a GETMAIN to expand the WEB pool. Time to queue the WEB can depend on the length of the queue and multiprocessor serialization and cache effects. If your z/OS system is running under a Hypervisor (LPAR or VM) using shared CPs, then there may also be logical processor dispatching effects (and these could be in the multi-millisecond range if the system is overutilized). Your code may be disabled from z/OS's point of view, but the logical processor it is running on is usually not disabled from the Hypervisor's point of view. Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
