FREEPOOL?
-- Shmuel (Seymour J.) Metz http://mason.gmu.edu/~smetz3 עַם יִשְׂרָאֵל חַי נֵ֣צַח יִשְׂרָאֵ֔ל לֹ֥א יְשַׁקֵּ֖ר ________________________________________ From: IBM Mainframe Discussion List <[email protected]> on behalf of Jack Zukt <[email protected]> Sent: Friday, October 4, 2024 4:43 AM To: [email protected] Subject: Re: Checking another address space private region ( Caution: This email did not originate from George Mason’s mail system. Do not click links or open attachments unless you recognize the sender and know the content is safe. Thank you all once again We suspect that the problem might be related to the open and close of multiple plain sequential files, a lot of them in MOD mode, but we can be completely off target. So, we are trying to have a look. It may be a bug but it also may be some not properly designed automation that we have built within netview. So far, using SMF and TDS we have been able to pinpoint two specific huge increase of storage use and we have sent that to the automation team. We are waiting for their feedback. Best wishes Jack On Thu, Oct 3, 2024, 21:01 Colin Paice < [email protected]> wrote: > Don't forget, the thread that causes an abend may not be the one causing > the problem. Thread 1 takes 99% of the storage. Thread 2 gets some > storage and abends. Looking in the systrace for the address space may show > a pattern of getting storage, but may not > > On Thu, Oct 3, 2024, 19:10 Michael Oujesky <[email protected]> wrote: > > > What type of files? VSAM, physical sequential, ZFS? SMF records > > ought to be able to tell. If PS, note that the buffer pool has to be > > freed after close. > > > > Michael > > > > > > At 11:40 AM 10/3/2024, Jack Zukt wrote: > > > > >Hi Peter > > >Thank you for your input. > > >We will look into it. > > >What we are trying to see is what is going on on a netview address space > > >that sometimes runs out of memory and terminates with an abend. We > suspect > > >that it may be related to open and close of files but we need a way to > > >check for it. > > >Best wishes > > >Jack > > > > > >On Thu, Oct 3, 2024, 13:36 Peter Relson <[email protected]> wrote: > > > > > > > The real question is what might you be thinking of looking at? > > > > > > > > SMF records are not going to help you find out what the problem is > > because > > > > they do not get down to the workunit level. > > > > > > > > Analysis of an address space is best done from a dump of that address > > > > space. Looking around with BLSACTV (if suitably authorized) may > easily > > run > > > > aground due to the fact that the data you are looking at is not > > necessarily > > > > in synch (because the data can be changing as it is retrieved). > > > > > > > > Peter Relson > > > > z/OS Core Technology Design > > > > > > > > > > > > > ---------------------------------------------------------------------- > > > > For IBM-MAIN subscribe / signoff / archive access instructions, > > > > send email to [email protected] with the message: INFO > IBM-MAIN > > > > > > > > > >---------------------------------------------------------------------- > > >For IBM-MAIN subscribe / signoff / archive access instructions, > > >send email to [email protected] with the message: INFO IBM-MAIN > > > > ---------------------------------------------------------------------- > > For IBM-MAIN subscribe / signoff / archive access instructions, > > send email to [email protected] with the message: INFO IBM-MAIN > > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
