Melinda,
I had a problem a while ago trying to restore a Netware 3.12 server.
The ADSM client on the Netware side would hang during a restore of
the server. Not only would the ADSM client hang but it would hang
the entire server. The restore we were running was being performed
in an attempt to restore the data from a server that was redirecting
blocks quite heavily and eventually packed it in. We were fortunate,
we were able to manually copy the files over to the new server and
update all the rights manually. The Novell support person here at
BCIT was not impressed.
When I originally opened the problem with ADSM support it was at
priority 2. I received one call back that asked the basic "what
version are you running type questions". After waiting for some
suggestions for a week or so and not getting a call or anything I
requeued my problem. I got a call back suggesting we do a trace. No
problem except when the Novell server hangs the trace file ended up
without the last bit of the trace. I was told the trace was not good
without the last piece of the trace. No new suggestions. Time went
on and about a week later the server with the redirects was really
getting ill. I raised the priority to 1 and waited for a call back.
I heard nothing for about 10 days. My manager asked for a status on
the problem and I sat there with my mouth open saying nothing. He
called our local IBM manager and I go a call back the next day from
ADSM support. The suggestion was to get another trace. After
forwarding that to ADSM support I heard nothing again. After another
week or so I requeued the problem. By this time we had rebuild the
drive manually. When I got a call back, the suggestion was to do a
TCP/IP trace from our VM system and also that we drop the problem to
priority 2 from priority 1 since the drive was now okay.
I have not done the TCP/IP trace because I do not have a server
available to mount a 9 GB drive on to try the restore. Prior to this
problem I had all the confidence in ADSM. Now I sweat a bit when we
have to restore a large amount of data.
Peter
> We are a moderately large ADSM installation, 3 servers, 3000 clients,
> 32G of database space.
>
> Last January, one of our servers began crashing three times each evening,
> apparently as the result of a corrupted database. I opened an incident,
> but got no callback until I called again and insisted that somebody call
> me. The person who finally called had me send log files covering a
> couple of evenings. Since then, I've heard nothing at all from them.
>
> The server continued to crash three times every evening, so many of
> the backups were not completing. That led me ultimately to raise the
> severity of my incident to 1, because the server was essentially
> unusable.
>
> Finally, in March, I figured out which client must be setting off the
> abends and moved it to a different server. That stopped the crashes,
> but nobody from IBM knows that, because the last time they contacted
> me was before then.
>
> And, of course, this is hardly a fix for the problem, because I'm left
> with a server that seems to have a corrupted database. I don't dare do
> anything to it, such as move to a new version or delete the bad client's
> files, for fear of making things worse. And I certainly don't know that
> I'd be able to restore that database from a current backup if I should
> need to.
>
> I really need to be able to use that server for that particular client,
> because it is the only one of the servers that has a tape robot that is
> in a different building from the client. In other words, I am no longer
> getting nightly offsite backups of that client (a large and important
> Novell server).
>
> I keep calling the Support Center and asking for a callback, but nobody
> ever calls me back. I have to assume that they aren't working on the
> problem, since they've never asked for a dump (I have *lots* of dumps).
>
> A few months ago, when the SHARE ADSM rep asked me to prepare an ADSM
> talk for SHARE next month, I suggested that that would be unwise from
> their point of view, as I would have no choice but to talk about their
> complete lack of support on this problem. She said the support manager
> would get in touch with me, but he/she has not yet done so.
>
> When I called the Support Center again last week to ask again for a
> callback, I was told that although my incident was a Severity 1, its
> priority was 3. If a server that crashes three times every evening is
> a priority 3, what is a priority 1? If anybody ever calls me back, I'll
> ask about that.
>
> I have been supporting IBM software for almost 30 years, and I've never
> had an experience like this. Indeed, this is orders of magnitude worse
> than the worst experiences I've ever had with any other IBM product.
>
> I apologize for airing dirty laundry in public, but I have long since
> exhausted all of the normal means for communicating with ADSM "Support",
> so I thought I'd give this a try.
>
> Melinda Varian,
> Office of Computing and Information Technology,
> Princeton University
>
>
|
