If a PTF has a ++DELETE, it will contain ++HOLD(sysmod-id) ... REASON(DELETE).... etc. You cannot restore those PTFs. The statement will look like ++DELETE(lmodname)
That is not to be confused with PTFs that delete MOD, MAC, or Data Element entries. Those would look like ++PNLENU(ABC) DELETE SYSLIB(sssss) DISTLIB(ddddd) . You can restore sysmods as long as they don't contain the ++DELETE MCS, but there are some considerations. 1. If you have newer PTFs that contain a PRE for the PTF that you are trying to restore, you will need to add those newer PTFs to your restore command, or even easier, just add GROUP to the RESTORE command. 2. If the sysmod that you are restoring has prereqs that you did not accept, then you will need to add those to the RESTORE command and re-apply them later. If you can accept the prereqs, then it will be a lot easier. Outside of that, it is up to you to set your policy of when you accept PTFs. The one policy to avoid is "NEVER", which makes it harder to back out fixes when you need to. Some of our CA products have releases that exist across multiple releases of CA, but need compatibility PTFs. Using a "never accept" policy on some of those products would allow you to restore back to levels that work with unsupported z/OS release. If you anticipate the need to run our product with z/OS 1.8, then that would be how you could do it. Bypassing error holds is up to you as the systems programmer. That is one of the great things about getting individual fixes instead of an unavoidable "push". For example, if we had a problem with our PDS Compression procedure, but you only use Backup and Restore, then it would be safe for you to bypass the error hold. You would be in control by analyzing the situation and making the decision, but I would definitely not advise a blanket BYPASS HOLDERROR technique. On Wed, Jan 30, 2019 at 5:34 PM Paul Gilmartin < [email protected]> wrote: > On Wed, 30 Jan 2019 23:00:09 +0000, Jesse 1 Robinson < > [email protected]> wrote: > > >I'm also curious about 'not applying', but 'not restoring' is a long > standing characteristic. A PTF may contain a ++DELETE among its effects. > There's a special system hold category for this bad boy that tells you it > cannot be restored because a key element is gone. They're fairly rare, but > they do exist. > > > It doesn't have to be that way. RESTORE ought to be possible if all the > parts to > rebulid that key element remain in the GLOBAL zone and the parts list > remains in > the CSI. SMP/E simply is not designed to do it. VMSES/E does better > because it > is not hindered by depending on ACCEPT and DLIB contents. > > -- gil > > ---------------------------------------------------------------------- > 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
