Re: LIC Upgrade

rob@xxxxxxxxx wrote:

I don't remember the exact errors I got. I even tried searching
to see if I had an old pmr on this. However if you plan on
upgrading to 545 from 540 or to 611 from 610 one had better
follow the book and eschew what Chuck is saying (even though his
advice is usually golden).

What I have suggested is only that the LIC-only install *should* be functional according to the design of the system and the way the [manner in which] PTFs for OS should be created and delivered. That is, I want to be clear that I am neither advising nor suggesting that anyone *should* "try this at home" ;-) Although... what harm would come of someone trying [if already using a manual install anyway], other than encountering an SRC, from which presumably a simple IPL up to DST then allows the start of the [since diagnosed by SRC as a required] OS install.?

FWiW: If such a requirement exists, then a possible origin can be fathomed by a test; albeit pointless to do so except as proof of origin for the requirement. By installing the OS "resave" level with the mystery PTF(s) that apparently are required to come from the install media versus coming from cumulative PTF media & having been applied only after a LIC-only install, if a later LIC-only install is allowed without the SRC, then there is some PTF by either function or mere existence allows the LIC install to complete without an SRC. And if that is possible solely by the existence of the PTF versus any function performed by that PTF during the install itself, then probably just applying some "marker PTF" would suffice; problem there however, is that a marker PTF would not be available on a cumulative since it merely indicates the REsave level for which the actual installation from that level is required [to effect the appearance of its being applied].

Chuck,
If you want to know bad enough, just try upgrading some system
from 610 to 611 or from 540 to 545 and see what you get if you
try it with just the LIC CD.

I'll get right on that ;-) ... after I win a lottery; although I must first get past the "must enter to win" issue. The reason I am asking, is that I have *no* ability to do that.

Maybe there's a technical reason they do it, like the same reason
they have respins of OS on occasion. Or maybe IBM just insists
on it. Maybe it's because I always do it with image catalogs and
something about the verify process is picky.

I am actually still very familiar with the technical aspects and the design [I have a good memory for some real useless stuff]; plus the difference between a "resave" and a "respin" [even more useless stuff, that only frustrates me for the terms incorrectly being used interchangeably :-Q]. I am of the opinion that if the development management and teams are so removed from the original design points that they would create & deliver a product which disallows the simplicity of performing a LIC-only install of the point release, then the system has become even _less_ of a "system" than ever. At least one highly negative implication is that use of an older SAVSYS or distribution\DSLO media of the same VRM for the OS would also be problematic to install from, as part of recovery of the OS-only... but if that were not problematic, then I am back to wondering why there would be any enforcement preventing the LIC-only install; i.e. they will protect one path by a restriction, only to leave another path dead in the water -- OK, so the likelihood of having to revert to or even inadvertently installing from older media is rare, but it does happen.

If the OS is a required part of the install, I wonder why QGPL, QUSRSYS, & XTND would be required also; or are they? And if they really are required, then why the noted documentation seems not to mention that little tidbit; i.e. only installing the OS & PTFs?

Regards, Chuck