I've been working on the ETR with our L2 support and waiting for some doc to confirm my assumptions. But I figure I better update this thread since so many seem to be speculating. I believe what happened is this.
One of the things DFSORT looks at is the values returned from SYSEVENT STGTEST. Some time back, STGTEST was updated so that the B and C values returned took into account available aux slots. Here is the text from APAR OA20116 Additionally the logic in IRAEVREQ ensures that the value returned in words 2 and 3 do not drive the system into an auxiliary storage shortage. The value returned in words 2 and 3 will only fill the AUX subsystem up to 50%. Examples: If the AUX subsystem is filled by 25% and the return value 1 contains 1000 frames. Then the return value 2 and 3 are set to 1000 frames + 25% of the AUX slots. >From the RMF I looked at, it appears when the page data sets were moved to the new dasd, they were also increased. This led to an increase in the value returned from STGTEST which led to DFSORT allocating more storage for sorting in memory. If you change the DFSORT default to EXPOLD=0, DFSORT will only factor in the A-value returned from STGTEST making it less likely to cause page stealing from other workloads. But again, I want to stress that we're still waiting for additional doc to verify all this so I don't want to encourage further speculation. I will update this thread when we confirm things. Have a nice day, Dave Betten DFSMS Performance Engineer IBM Corporation email: bet...@us.ibm.com DFSORT/MVSontheweb at http://www.ibm.com/storage/dfsort/ IBM Mainframe Discussion List <IBM-MAIN@listserv.ua.edu> wrote on 03/01/2013 07:23:30 AM: > From: "Adams, Anne (DTI)" <anne.ad...@state.de.us> > To: IBM-MAIN@listserv.ua.edu, > Date: 03/01/2013 08:23 AM > Subject: Re: DFSORT Weirdness > Sent by: IBM Mainframe Discussion List <IBM-MAIN@listserv.ua.edu> > > The LPAR has 14GB of memory and the six page data sets have (yikes!) > five different sizes (1000, 1000, 1500, 2000, 2500, and 3000 > cylinders). Apparently the statement, we didn't change anything, > really meant, we didn't change anything that should affect you. > > Anne R. Adams > DTI, Systems Engineering > State of Delaware > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU > ] On Behalf Of Vernooij, CP - SPLXM > Sent: Friday, March 01, 2013 2:17 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT Weirdness > > Anne, > > Just for the picture: how much memory do you have in your LPAR and > what is the total size of the page datasets.? > > Kees. > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU > ] On Behalf Of Adams, Anne (DTI) > Sent: Thursday, February 28, 2013 18:29 > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT Weirdness > > Ok - this is EXACTLY what we're seeing now. The one issue is that > when the page data sets fill up, they never get emptied and we end > up having to add new ones. What do people do? We can't IPL every > week or go on adding page datasets indefinitely. > > Anne R. Adams > DTI, Systems Engineering > (302) 298 - 3196 > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU > ] On Behalf Of Vernooij, CP - SPLXM > Sent: Thursday, February 28, 2013 10:59 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT Weirdness > > I have seen DB2 utilities with large sorts (reorgs of partitioned > table spaces, which does a lot in parallel), that consumed 6 - 8 GB > memory of the 16GB in the LPAR. Nobody was really harmed by it, only > you could see RMS paging out heavily old stuff to make room for the > utilty. On these moments, you should have a page configuration ready > for these page outs. There was no heavy page ins, so performance was > not harmed. This is real memory management of modern, large > machines, which should not be undervalued (correct English?) > > Kees. > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU > ] On Behalf Of Elardus Engelbrecht > Sent: Thursday, February 28, 2013 16:51 > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT Weirdness > > Kees Vernooij wrote: > > >Again: the fact that DFsort acts differently on the new device > could be caused by CFW. With CFW dfsort could decide to use SORTWKs, > without CFW DFSORT could decide that this is too slow and use memory. > > Very true! Will the DFSORT gurus answer on your statement this or > next month? ;-D > > I'm watching this thread, because my own DFSORT + ICETOOL jobs are > grabbing and eating everything - CPU, Inits, Memory and DASD space, > name it, like there is no tomorrow... ;-D > > Groete / Greetings > Elardus Engelbrecht > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send > email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > ******************************************************** > For information, services and offers, please visit our web site: > http://www.klm.com. This e-mail and any attachment may contain > confidential and privileged material intended for the addressee > only. If you are not the addressee, you are notified that no part of > the e-mail or any attachment may be disclosed, copied or > distributed, and that any other action related to this e-mail or > attachment is strictly prohibited, and may be unlawful. If you have > received this e-mail by error, please notify the sender immediately > by return e-mail, and delete this message. > > Koninklijke Luchtvaart Maatschappij NV (KLM), its subsidiaries and/ > or its employees shall not be liable for the incorrect or incomplete > transmission of this e-mail or any attachments, nor responsible for > any delay in receipt. > Koninklijke Luchtvaart Maatschappij N.V. (also known as KLM Royal > Dutch Airlines) is registered in Amstelveen, The Netherlands, with > registered number 33014286 > ******************************************************** > > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send > email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send > email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > ******************************************************** > For information, services and offers, please visit our web site: > http://www.klm.com. This e-mail and any attachment may contain > confidential and privileged material intended for the addressee > only. If you are not the addressee, you are notified that no part of > the e-mail or any attachment may be disclosed, copied or > distributed, and that any other action related to this e-mail or > attachment is strictly prohibited, and may be unlawful. If you have > received this e-mail by error, please notify the sender immediately > by return e-mail, and delete this message. > > Koninklijke Luchtvaart Maatschappij NV (KLM), its subsidiaries and/ > or its employees shall not be liable for the incorrect or incomplete > transmission of this e-mail or any attachments, nor responsible for > any delay in receipt. > Koninklijke Luchtvaart Maatschappij N.V. (also known as KLM Royal > Dutch Airlines) is registered in Amstelveen, The Netherlands, with > registered number 33014286 > ******************************************************** > > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send > email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
