Re: Restoring MSCS cluster node
2002-08-02 18:04:25
Passive/Active
It pays to perform the test and actually wipe out the c drive of one of the
nodes.
In testing we performed a C drive backup and a system state backup.
In the event SYSTEM A crashes with a system drive failure. SYSTEM B should
take over. Get SYSTEM A up, install tsm client, then restore the c$ drive.
Lastly, restore the system state. Should be able to fail over to SYSTEM A.
One thing we did find quicker was using a single system as a collector for
all system states of all cluster nodes. Came to about 250-300 MB per system.
-----Original Message-----
From: Jim Smith [mailto:smithjp AT US.IBM DOT COM]
Sent: Friday, August 02, 2002 4:21 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Restoring MSCS cluster node
J.P. (Jim) Smith
TSM Client Development
smithjp AT us.ibm DOT com
Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
cc:
Subject: Restoring MSCS cluster node
Hi.
Has anyone got an idea how to perform "bare metal" restore of MSCS
cluster node (win2k, winNT)? I have performed a lot of tries to do that.
Sometimes it succeeds sometimes doesn't. According to the redbook
sg24-6141 "Deploying the TSM client in a win2k env" I restored ALL
system object including cluster db but cluster services were not running
(of course) and restore failed. So I restored system objects except
cluster db (first time it succeeded).
If you know any better redbook than above or if you have any experience
in it please HELP ME!
--
--== KKus ==--
|
|
|