Andrew, I am not asking about batch jobs that can start long-lasting or somehow disconnected Unix processes that outlive the batch job execution. I am only asking about synchronous processes started and completed in the course of one batch job. Your reference to “sshd” is what has confused me. What does the “ssh” demon process have to do with a simple batch job that runs shell commands and a python (or go, etc.) program for some application-specific reasons?
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). And again, this is all out of pure curiosity – there is zero chance I can ever get our local IEFACTRT routine or end-of-step processes changed to do any such reporting. Peter From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> On Behalf Of Andrew Rowley Sent: Wednesday, June 26, 2024 8:19 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: CPU and I/O statistics for BPXBATCH executions? On 26/06/2024 10:05 pm, Seymour J Metz wrote: > 1. If the installation chooses to run multiple processes in a single AS, > I know of no way to separate the data for individual processes. Is > there a new SMF record type for that purpose? There is some data for individual processes in the type 30 record in the z/OS Unix Process Section. However, because you can have multiple processes, the information can be spread across multiple records if there is too much information for a single record. So it isn't necessarily in the same SMF record as e.g. the Processor Accounting section. > 2. The converse issue is aggregating all of data for processes > spawned by a single job. Is there anything in the type 30 to > identify the original job? The issue of which processes should be aggregated is also difficult. For example, you probably want processes spawned by sshd accounted separately (although there are multiple levels, some belong to sshd, some to the user logging in.) But if a batch job does some work under UID 0, it still belongs to the batch job. There is a session id which identifies related processes, but the original job might end long before or long after the other processes so it can be difficult to relate them in the SMF data. -- This message and any attachments are intended only for the use of the addressee and may contain information that is privileged and confidential. If the reader of the message is not the intended recipient or an authorized representative of the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by e-mail and delete the message and any attachments from your system. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
