DB2 "upgrade"

Roger Miller

DB2 "upgrade"
On Tue, 24 Dec 2002 11:05:20 -0800, Taddei, Cathy
<[login to unmask email]> wrote:

>Now I'm curious. I did not know IBM provided for mass maintenance for DB2
>in this way. How do they inform you of changes you need to make to your
>catalog? For example, looking at the last round of mass maintenance that I
>did on DB2 V6: UQ55014 included an ACTION HOLD to remove orphaned rows
>SYSTABAUTH. There were several other PTF's with SPUFI's to detect bad
>values in SYSTABLES, SYSROUTINES, and SYSTABLESPACE, and unique methods to
>clean up any problems found. UQ47800 instructed me to run job DSNTEJ6W to
>create the WLM_REFRESH stored procedure, which was new with that PTF.
>IBM have created a package of maintenance for me with instructions to do
>those things? If so, I would be very interested in that. I would never
>have to read holddata again!
>On the other hand, I have been in shops where I had to apply several years
>worth of maintenance, and there was nothing dangerous or unusual about it.
>You can break it up into smaller chunks if you want, but it's basically the
>same work and the same result. Plus, you know you've done every HOLD
>and read every HOLD DOC, so you have confidence in the integrity of your
>system and are up to speed on the new stuff. I think it's the safest way
>go no matter how far you are behind. Or am I missing out on something? If
>IBM offers actual "maintenance only upgrades", I'd be very interested!
>Cathy Taddei
>-----Original Message-----
>From: Ed Long [mailto:[login to unmask email]
>Sent: Tuesday, December 24, 2002 6:48 AM
>To: [login to unmask email]
>Subject: Re: DB2 "upgrade"
>The judgement part of our job comes into play here.
>The key decision points are, how back level is the
>starting point? How much maintenance is missing?
>I interpreted the initial email as saying that they
>were very backlevel but on the same release. In that
>instance, an RSU or full replacement, would be easier
>and ultimately safer since the new system would have
>been tested by IBM before release.
>The reason I mentioned SDSNSAMP in my response was to
>point out that rerunning ZPARMS and rebuilding the WZP
>type SP's is now (v7) a pretty common bit of HoldData.
>It perhaps wasn't as clear as it could have been.
>The heart of the matter remains testing the upgraded
>system before exposing production to it, regardless of
>how you get there.
>This email is confidential and may be legally privileged.
>It is intended solely for the addressee. Access to this email by anyone
else, unless expressly approved by the sender or an authorized addressee,
is unauthorized.
>If you are not the intended recipient, any disclosure, copying,
distribution or any action omitted or taken in reliance on it, is
prohibited and may be unlawful. If you believe that you have received this
email in error, please contact the sender, delete this e-mail and destroy
all copies.

What I expected most customers to do was to apply service that matches the
RSU levels, so that the service levels are consistent across our software
stack. These are still just a combination of service, and there is not any
technique that can avoid reading the holddata and performing the needed
actions. We are seeing more customers who can only upgrade software a
couple (one to four) times a year, so one product at a time does not do the
job. This technique does not eliminate the need to test either, especially
where your usage is unique.

The other significant use of the CST process is to provide a recommendation
for a service level when customers migrate to a new release or version.

This is my opinion, not necessarily that of my employer.

Roger Miller