On May 6, 2007, at 10:57 PM, Brian Westerman wrote:
-------------------SNIP----------------------
There will be exits that need to be re-written, but even if you had
stayed
within the N+/-2 scheme, you still would have run into the issue a
some
point. Sites that are fairly vanilla (home grown and exit-wise),
don't
really have much to fear from making the big jumps. There are a
lot of
people who got bitten just from the JES2 changes from z/OS 1.6 to 1.7.
So, while there will be issues that need to be addressed, they are
the same
ones that would have to be addressed if the original thread creator
was to
jump in multiple steps, (OS/390 to z/OS 1.4, then 1.6 then 1.8),
they just
would have had a great deal of extra conversion work that might
have masked
the basic changes that they needed to make to their own "home
grown" code.
I'm not into making people jump through hoops to get from point A
to Z, (get
it?), if I can get them there directly, why force them to go
through B, C D,
etc. when all they will really get from it is extra work?
==========SNIP++++++++++++
I used to work at a "vanilla" shop a while ago. We had an occasion to
put on several years of maintenance at one time because of management
sitting on their collective asses and not upgrading the computer.
As you say exits have to be looked at (re-written) etc and those were
the types of things that we ran into and all had been anticipated.
SInce the exits were simple it was pretty much a stream lined
operation. I think the upgrade of OEM products proved to be a PITA
more than the exits. One of larger issues we had with OEM is that at
that time the (some) didn't have a clue as to what release of their
software would run with each IBM release. I gave the project to a jr
system programmer and after a week he came back with a report that
made my spine cringle. I then decided to do the project myself and
came away almost crying as the vendors just DID NOT have a clue. I
had to set up a conference call with each one of the vendors and make
sure that people that were supposed to be there on the conference
line were there. This was approximately 10 years ago (IIRC before the
N+2 which I think all the vendors loved as they could get their arms
around it).
Having said that we tested everything that we could production (not
every job just a cross section) and everything worked as expected.
The switch over happened and almost zero issues for the first say 4
hours then one job JCL'd then 2 then 3 then 4 after about 30 minutes
we had 8 jobs that had jcl'd. Out of 200+ jobs we had 8 jobs that
failed. It turned out that there was a JCL DD DLM= that did us in Jes
2 changed the rules and it wasn'[t terribly clear if it was the
converter interpreter or JES. I sat down and skimmed the cover
letters and found out that JES 2 was the "culprit". I was able to
convey to the production control people what had to be looked for and
how to change it.
I cannot blame OEM I can't blame exits just JES and why there didn't
flag the ptf as needing to be special handling.
In my case it was IBM PTF and OEM vendor(s) that did not understand
the IBM level OS issue.
Ed
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
