I would put the TSM scheduler into cluster administrator, because that
essentially stops/starts the scheduler service for you when you do a
failover.  Follow the instructions in one of the appendixes of the client
manual.  Should take care of the problem.

______________________________
John Monahan
Senior Consultant Enterprise Solutions
Computech Resources, Inc.
Office: 952-833-0930 ext 109
Cell: 952-221-6938
http://www.compures.com




             "Ochs, Duane"
             <Duane.Ochs AT QG DOT CO
             M>                                                         To
             Sent by: "ADSM:           ADSM-L AT VM.MARIST DOT EDU
             Dist Stor                                                  cc
             Manager"
             <[email protected]                                     Subject
             .EDU>                     Re: Client : MS Cluster active -
                                       active

             10/13/2003 02:30
             PM


             Please respond to
             "ADSM: Dist Stor
                 Manager"
             <[email protected]
                   .EDU>






Each time it has happened both systems eventually were rebooted and had the
same results. 1) Blaster virus 2) security patches and 3) SP update.

It occured each time and I do not always get notified when our Intel
servers
group performs these functions.

We do not have the two schedulers configured as cluster services, but TSM
is
aware that they are a clustered system.

I reviewed the dsmsched.log to find that only 4 files were scanned and 4
were backed up. Out of 1.4 million.


Duane


-----Original Message-----
From: John Monahan [mailto:JMonahan AT COMPURES DOT COM]
Sent: Monday, October 13, 2003 2:17 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Client : MS Cluster active - active


"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU> wrote on 10/13/2003
09:44:48 AM:

> TSM Server 5.1.6.2 ( AIX)
> TSM Client 5.1.5.15 (W2K cluster)
>
> Cluster node a - Has drive f$
> Cluster node b - Has drive g$
>
> To apply patches to the systems, nodea f$ is failed to node b, patches
> applied to nodea and rebooted, f$ and g$ are failed to node a, Node b
then
> gets patch and rebooted then g$ is failed back to node b.
>
> At this point, backups can not be performed on either f$ or g$, only the
top
> level directories are backed up and no failures are indicated in the
Client
> log or the TSM server activity log.

Only top level directories are backed up on f: and g:?  How do you know
this?


> Each node has a scheduler, If I restart the schedulers the scheduled
backups
> run fine at next schedule.

What schedmode are you using?  Is your scheduler service defined in cluster
administrator?

What if you do a failover without doing the patches part, does the same
thing happen?