My last several Serverpac installs had very few issues. Most were self-inflicted. Like the time I failed to order the optional regulated encryption. Or this last time, when do to poor timing, I had to start with a z/OS 2.3 archive, then add Java 😊
For at least the last 3, implementation in production was a non-event. Which is as it should be. > -----Original Message----- > From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> On > Behalf Of Tom Brennan > Sent: Thursday, July 22, 2021 7:38 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: Serverpac installs January 2022 and beyond - Requests > > "seems everyone has a better way" > > I think you hit on the root of the problem. With Windows and Linux > installs, everyone (generally) does things the same exact way, including > filenames and directory locations. They don't have the problems we have > with mainframe installs. > > On 7/22/2021 7:19 AM, Carmen Vitullo wrote: > > I think I IPL'd the CPAC system that was built from the ServerPac once > > in my career, it was the company/dept's standard and we had a small LPAR > > built just for that reason. Documentation was provided, IPLing the CPAC > > system was only done to proceed with the ServerPac install. > > > >  moving on from that company I moved to a different process, seems > > everyone has a better way, for me building the target sysres and zfs > > file systems, running some IVP tests and build my new master catalog, > > and IPL that system on my sandbox system. > > > > I have a documented process to copy/migrate the new version or maint to > > production that works well even for someone who's not a z/os sysprog. > > > > Carmen > > > > > > On 7/21/2021 1:19 PM, Tom Brennan wrote: > >> Same with me when I ran ServerPac installs - I never IPL'd using the > >> datasets provided by the installer such as catalogs, RACF, spool, SMF, > >> page, etc. I never understood IBM's reason for doing that, and also > >> never understood the reason for running the system validation jobs on > >> the vanilla system. What was much more important for us was IPLing > >> the new res pack on a sandbox system with our own system datasets, > >> parms, and usermods - and then solve any issues that may come up. > >> > >> So those IBM-supplied system datasets were never used, and although I > >> could not delete them using the CPP dialog, I would always set them to > >> 1 track or 1 cylinder before running the alloc job - just to save space. > >> > >> It just made little sense to me to prove the vanilla system from IBM > >> works correctly. Of course it does, otherwise why would they send it > >> to me? > >> > >> On 7/20/2021 10:23 PM, Gibney, Dave wrote: > >>> I don't know how it would work with zOSMF, but I don't worry about > >>> the dataset sizes of my SMPE target datasets. Because I never IPL > >>> using them. > >>> I copy to new SYSRES, FDR and ADRDSSU dataset copies to single > >>> extents. Of course, I rarely (maybe 5 to 5 times in 30 years) put > >>> maintenance into a running system > >> > >> ---------------------------------------------------------------------- > >> 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
