The JES2 checkpoint is the only mechanism that maps spool data. Assuming two outboard copies of the checkpoint--wherever located--either one can be reconstructed dynamically from the other even at IPL time. If both are lost or destroyed--wherever located--you are SOL at IPL time. In particular, the checkpoint *cannot* be recovered or reconstructed from spool data alone. Without a usable copy of the checkpoint somewhere (in memory counts), spool data is meaningless. Cold start time.
So never put both primary and alternate copies in CFs in a single CEC. It's doable with dual CECs but maybe not a great use of CF memory. There are two chief advantages of CF checkpoint: performance in general (responsiveness) and performance balancing among CECs of widely different speeds (XCF tends to serialize accesses). If you choose to use CF for checkpoint in a single CEC, put *only* the primary in a structure. . . J.O.Skip Robinson Southern California Edison Company Electric Dragon Team Paddler SHARE MVS Program Co-Manager 323-715-0595 Mobile 626-543-6132 Office ⇐=== NEW robin...@sce.com -----Original Message----- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Vernooij, Kees (ITOPT1) - KLM Sent: Monday, July 31, 2017 5:34 AM To: IBM-MAIN@LISTSERV.UA.EDU Subject: (External):Re: JES2 Checkpoint question I think I remember from the past, when we had CKPT in the CF, that only 1 of every 10 writes to the primary ckpt is done to the secondary. So having the secondary on dasd will cause minimal performance degradation, if that were a concern. Kees. > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] > On Behalf Of Carmen Vitullo > Sent: 31 July, 2017 14:24 > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: JES2 Checkpoint question > > That what I was thinking, according to the doc, and its not 100% clear > I can recover from a lost CKPT, so does that assume just one lost CKPT > while JES2 is up, or during a shutdown, there are so many different > recovery scenarios' but none that fit my situation. > It was my bad planning, putting both CKPT's in the CF's and not > planning this thru to the power down - POR lessons learned here. > thank again all > Carmen > > > ----- Original Message ----- > > From: "Kees Vernooij (ITOPT1) - KLM" <kees.verno...@klm.com> > To: IBM-MAIN@LISTSERV.UA.EDU > Sent: Monday, July 31, 2017 5:16:09 AM > Subject: Re: JES2 Checkpoint question > > I agree, that I can recover JES2 to use new checkpoints if the > originals are definitely gone, but that was not the case here. > What can I recover, when both checkpoints have been emptied by the POR? > > Kees. > > > -----Original Message----- > > From: IBM Mainframe Discussion List > > [mailto:IBM-MAIN@LISTSERV.UA.EDU] > On > > Behalf Of Gerhard Adam > > Sent: 31 July, 2017 12:11 > > To: IBM-MAIN@LISTSERV.UA.EDU > > Subject: Re: JES2 Checkpoint question > > > > I disagree. The point of reconfiguration is also to recover from > > checkpoint loss. > > > > Sent from my iPhone > > > > > On Jul 31, 2017, at 3:00 AM, Vernooij, Kees (ITOPT1) - KLM > > <kees.verno...@klm.com> wrote: > > > > > > But still, if both checkpoints were in CF's, the POR cleared both > > > of > > them, so there is nothing to reconfigure. A COLD start would indeed > > be the fastest. > > > > > > Kees. > > > > > >> -----Original Message----- > > >> From: IBM Mainframe Discussion List [mailto:IBM- > m...@listserv.ua.edu] > > On > > >> Behalf Of Gerhard Adam > > >> Sent: 28 July, 2017 18:17 > > >> To: IBM-MAIN@LISTSERV.UA.EDU > > >> Subject: Re: JES2 Checkpoint question > > >> > > >> I wasn't sure if you were aware that you can make up a NEWCKPT > > >> dynamically and if it doesn't exist, JES2 will prompt you to > > >> create > > it. > > >> > > >> That way if you don't have an existing definition, you can still > get > > a > > >> new data set created and migrated to. > > >> > > >> > > >> > > >> -----Original Message----- > > >> From: IBM Mainframe Discussion List [mailto:IBM- > m...@listserv.ua.edu] > > On > > >> Behalf Of Carmen Vitullo > > >> Sent: Friday, July 28, 2017 7:15 AM > > >> To: IBM-MAIN@LISTSERV.UA.EDU > > >> Subject: Re: JES2 Checkpoint question > > >> > > >> thank you > > >> I decided to do a cold start due to the fact while an attempt to > move > > to > > >> NEWCKPRT1 or 2 during startup I received I/O errors on the CKPT > > files, > > >> due to time constrains on my downtime I had to get the systems > > >> back > > up. > > >> I've since read there are somethings I could have tried to > > >> recover, > > but > > >> at the time....there was no time. > > >> > > >> > > >> Carmen > > >> > > >> ----- Original Message ----- > > >> > > >> From: "Gerhard Adam" <gada...@charter.net> > > >> To: IBM-MAIN@LISTSERV.UA.EDU > > >> Sent: Friday, July 28, 2017 9:08:21 AM > > >> Subject: Re: JES2 Checkpoint question > > >> > > >> You shouldn't need both on DASD. However, I am curious as to why > you > > had > > >> to COLD start. > > >> > > >> > > >> > > >> -----Original Message----- > > >> From: IBM Mainframe Discussion List [mailto:IBM- > m...@listserv.ua.edu] > > On > > >> Behalf Of Carmen Vitullo > > >> Sent: Friday, July 28, 2017 6:55 AM > > >> To: IBM-MAIN@LISTSERV.UA.EDU > > >> Subject: JES2 Checkpoint question > > >> > > >> I've looked and looked at the Init and Tuning REF and Guide and I > > can't > > >> seem to find an answer to my question, I'm hoping someone can > provide > > a > > >> sanity check for me. > > >> I have a 2 system JES2 MAS both running z/OS 2.1, both > > >> checkpoints > > are > > >> currently in a CF with mode=duplex and duplex=on. > > >> my issue is both CF's are in the same CEC so a POR of the box for > > maint > > >> last month I lost all my prod spool data and was forced to > > >> perform > a > > >> JES2 cold start. I need to move CKPT1 to DASD and I want to leave > > CKPT2 > > >> in the CF, so my question is can I still run with mode=duplex and > > >> duplex=on in that configuration, or do I need to have both CKPT's > on > > >> DASD. > > >> Any insight or recommendations would be very helpful, thank you > > >> all ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
