I would like to investigate this issue here, but I don't know how to find a DDDEF that might point to the /etc or /var file system. As I said previously, our /etc and /var file systems in production are 'permanent' in that they are not updated by our maintenance migration. But I am concerned about the situation on the (only) system where we actually run SMPE.
. . 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 [email protected] -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Art Gutowski Sent: Friday, February 09, 2018 9:53 AM To: [email protected] Subject: (External):Re: Best Practices for z/OS Maintenance >We don't mount the IBM provided /etc /var file systems for actual use. We >use those to compare with what we have. >AFAIK, there is NOT any SMPE that updates those anyway, its just what >Serverpac provides. >I never mount those filesystems for SMPE maintenance and never get any errors. SMP/E LIST DDDEF will confirm that assumption. It's been a while since I've had to do z/OS maintenance, but I've learned not to take anything for granted with SMP/E, and especially with z/OS UNIX filesystems. IF you find any DDDEFs pointing to /etc or /var, they would need to be changed to some variation of /service as you hopefully do with your version FS. If you manage multiple target and distribution pairs, I highly recommend using automount to ensure SMP/E mounts the matching version root FS for the target zone and SYSRES. A colleague (who frequents this list), set this up at a prior shop of ours. It was not trivial. He had a ServiceLink Q&A open with IBM for weeks, and there was much discussion among the team and extensive testing. However, once the various maps were defined (we had to support multiple versions of languages as well), the DDDEFs were updated, and we modified our cloning process to keep up, everything worked flawlessly. IJS, I would not take the lack of errors as a golden stamp of approval. We were burned badly by this assumption - our SYSRES and version FS (and SMP/E) got out of sync, and we had no obvious indication. If memory serves, we had to dig through the SMP/E LOG (the job SYSOUT was probably long gone) to find that SMP/E had done exactly what it was told to do: apply UNIX elements to the wrong path/filesystem. Art ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
