You have until the end of today to submit a presentation proposal on this (z/OS Unix SMF 30 parsing) for GSE UK Annual Conference, Scott.
Not to put pressure on you in public… 😊 Cheers, Martin From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of Scott Chapman <000003fffd029d68-dmarc-requ...@listserv.ua.edu> Date: Friday, 28 June 2024 at 12:30 To: IBM-MAIN@LISTSERV.UA.EDU <IBM-MAIN@LISTSERV.UA.EDU> Subject: [EXTERNAL] Re: CPU and I/O statistics for BPXBATCH executions? Yeah, one might think. One might be surprised. I haven't looked at this for some time so I don't remember all the details. But I remember I wrote something about this at one point in time. Eventually I found an old CMG paper I wrote (looks like c. 2011) about running Java on z/OS and found the below bit. It seems you might want to try running under BPXBATSL to see if that impacts what's reported. If you have shell scripts calling other "unix" commands that the combination of BPXBATSL and the _BPX* settings here might (possibly) get you to the all in one accounting. Scott Chapman From 2011: SMF 30 and End-of-job Accounting The choice of BPXBATCH vs. BPXBATSL vs. JZOS directly impacts the SMF 30 records that are created for a Java batch job. Note that there are USS environment variables that may impact the details presented here. In particular, _BPX_BATCH_SPAWN, _BPX_SPAWN_SCRIPT, and _BPX_SHAREAS seem like they should impact the operation of BPXBATCH, but in my limited testing they didn’t. Again, testing in your own particular environment is highly suggested! The results presented here are believed to be valid for a standard or default configuration. Also note that BPXBATCH was using the SH option to run a shell script which can afford more space in the 100 byte JCL PARM, but causes some of the problems noted here. If you expect to run a single batch job with a single step and get a single SMF 30 step record (subtype 4) and a single SMF 30 job record (subtype 5), you may be surprised. In my testing that was only true when using BPXBATSL. When using BPXBATCH, I got 4 job records and 6 step records. When using JZOS, I got 2 job records and 2 step records. On Thu, 27 Jun 2024 00:54:34 +0000, Farley, Peter <peter.far...@broadridge.com> wrote: >For such a limited batch job, it ought to be possible to report all I/O’s, CPU >and memory statistics for all processes that were started and ended during the >job execution in one or two aggregated set of messages to the batch job >JESYSMSG and JESMSGLG outputs such that the combination reports all of the >actual usage, whether z/OS batch or Unix command(s). ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN Unless otherwise stated above: IBM United Kingdom Limited Registered in England and Wales with number 741598 Registered office: PO Box 41, North Harbour, Portsmouth, Hants. PO6 3AU ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
