I had this problem, but it was fixed by making sure that the box
titled "does affect resource group" was NOT checked when the generic
resource group was defined. Was this done?



At 10:49 AM 5/15/2002 -0400, you wrote:

Dave,

Thanks for your help yesterday.  Greatly appreciated.  All seemed to have
gone well... didn't test for failover as of yet (SA's had other things to
do).  One question if you don't mind...
It seems (according to the SA's) that after installing TSM and setting up
all of the services, the SA's are not able to move services from one
machine to another.  They end up having to reboot both
machines to free up the resources.  Is this anything that you've heard of
relating to TSM?  There is nothing that I see that indicates TSM is
causing this problem.  Moreover, I can't imagine how TSM
could be holding on to a resource and preventing it from moving.

Thanks again, Joe


-----Original Message-----
From: Dave Canan [mailto:ddcanan AT ATTGLOBAL DOT NET]
Sent: Tuesday, May 14, 2002 12:30 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Microsoft cluster server / TSM


         I recently set something like this up, and would like to discuss
it with you, as I see several issues the way you're doing it. Is there a
phone #  I could call you at? You can also e-mail me at
ddcanan AT attglobal DOT net if you want with the phone number.


At 09:37 AM 5/14/2002 -0400, you wrote:
>Need a little help.  I'm about to set up a TSM for Microsoft cluster
>server.  It will be an active/active configuration.  I plan on doing it as
>follows:
>
>
>Machinge A
>Physical disk H is a cluster group
>
>Machine B
>Physical disk I is a cluster group
>
>Physical disk Q (quorum) is a cluster group
>
>Machine A (normally owns disk H)
>install local scheduler... always running
>install scheduler for cluster group H (pointing to a dsm.opt file on the H
>drive)... always running
>install scheduler for cluster group Q (pointing to a dsm.opt file on the Q
>drive)... always running
>install failover scheduler pointing to the dsm.opt file on the I: drive
>which will manage cluster group I: in the event of failure (this will be a
>"Generic Service Resource for Failover" this
>scheduler is not on auto start=yes)
>
>Machine B (normally owns disk I:)
>install local scheduler... always running
>install scheduler for cluster group I (pointing to a dsm.opt file on the I
>drive)... always running
>install scheduler for cluster group Q (pointing to a dsm.opt file on the Q
>drive)... always running
>install failover scheduler pointing to the dsm.opt file on the H: drive
>which will manage cluster group H:in the event of Machine A
>failure.  (this will be a "Generic Service Resource for Failover"
>this scheduler is not on auto start=yes)
>
>
>Here are my two questions:
>Should the scheduler for Q be running on both machine A and machine B?
>
>For machine A...when I set up "Depencies" in cluster manager for the
>"Generic Service Resource for Failover" and it states to "add all physical
>disk resources", do I add both H and I?
>
>I can't seem to get a straight answer on this one.
>
>Regards, Joe

Money is not the root of all evil - full backups are.

Money is not the root of all evil - full backups are.