Re: Backup - revisited

[CRPence <CRPbottle@xxxxxxxxx>]

  That this was an upgrade from v5r2 to v5r4, so the restore was 
actually part of an upgrade scenario versus a standard DR, the origin of 
the problem in understanding this issue is probably now clear.  It seems 
that the discussion is not about DR, but about an _upgrade_ that is 
performed with _data migration_ from a prior release.  There are a 
variety of data migration strategies.  When an upgrade is thrown in, the 
best choice is based on the capabilities of the hardware support for the 
release levels involved.  Ideally the target system can be installed to 
the release level of the source system with only the LIC & OS, to which 
the restore-21 is performed, where the target is then upgraded from the 
lower to the higher release level; ideally, because that limits the 
number of release upgrades to just one.

  The most common failure in such upgrades is for an *assumption* that 
it can be done simply by restoring stuff, rather than following a well 
understood and tested migration path.  And for failure to follow the 
path with a good checklist, most typically the biggest problems are a 
direct result of improper migration of the user data in quasi-system 
user libraries, primarily QUSRSYS and QSYS2.  As such my guess is, that 
rather than anything properly termed metadata being at issue, the issue 
is user data.  Although it is called /user data/ because it was supplied 
by the user\operator to an application on the server, on another 
platform it might be called instead, /application data/; where in this 
case, the "application" is the OS.

Regards, Chuck

Ingvaldson, Scott wrote:
> Additional information from John:
>
> > When BBACKUP was used, ENDSBS OPTION(*IMMED) preceeded BBACKUP. The
> >  commands used by BBACKUP are:
> >
> > SAVSYS DEV(TAP35) ENDOPT(*LEAVE) SAVLIB LIB(*NONSYS) DEV(TAP35)
> >  ENDOPT(*LEAVE) ACCPTH(*YES)
> > SAVDLO DLO(*ALL) DEV(TAP35) ENDOPT(*LEAVE)
> > SAV DEV('/QSYS.LIB/TAP35.DEVD')
> >  OBJ(('/*') ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT) ('/QNTC' *OMIT))
> >  ENDOPT(*UNLOAD) UPDHST(*YES)
> >
> > The SA attempted to restore from 820 (?) to 520. I was not present 
> > when he did this. All he offered was that the metadata was not 
> > saved by the above commands. I asked what was included in the 
> > metadata, and got no answer. When I showed him a page from
> > infocenter detailing the processes used by option 21, he stated
> > that *something* is different.
>
> > Stated that he had been on the phone with IBM for weeks.
> >
> > Old box was v5r2 and new box is at v5r4. I have NO idea what was 
> > loaded on the new box.
> >
> > All of this is to do DR planning at some point. He even stated that
> > option 21 was also non functional. But he quickly moved on. This 
> > makes absolutely no sense to me.
> >
> > Does any of this help? Any additional information that I might be 
> > able to dig up that could be worthwhile?
> >
> > Our mutual supervisor has asked me to contact IBM support for 
> > clarification. I would dearly like the call to be worthwhile. But, 
> > I already see that I am missing a considerable level of detail. For
> > starters, what did he do, on what kind of system.