ADSM-L

Re: Perform backups on clustervolumes

2006-11-09 13:47:51
Subject: Re: Perform backups on clustervolumes
From: Andrew Ferris <aferris AT MRL.UBC DOT CA>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Thu, 9 Nov 2006 10:46:13 -0800
Hi Peter,

Your last question is probably solved by using the "NWEXITNLMP NO" option in 
your DSM.opt files. That means you don't have to press a key to make the TSM 
client close.

I can't help too much with cluster backups using "dsmc sched" because we use 
DSMCAD. It's been very stable for us. Here's a standard DSM.opt file for a 
cluster resource using DSMCAD:

*  Cluster Config for XXXX Volume
COMMMETHOD                TCPIP 
TCPSERVERADDRESS         xxx.mrl.ubc.ca
TCPPORT                  1500

NODENAME                  XXXX_VOL

HTTPPORT 1611
ERRORLOGRETENTION 7
SCHEDLOGRETENTION 7
MANAGEDSERVICES WEBCLIENT SCHEDULE
DOMAIN Volume:
CLUSTERNODE YES
PASSWORDDIR Volume:\Tivoli
PASSWORDACCESS GENERATE
NWPWFILE YES
ERRORLOGNAME Volume:\Tivoli\dsmerror.log
SCHEDLOGNAME Volume:\Tivoli\dsmsched.log
SCHEDMODE POLLING 
NWEXITNLMP NO
  
  We have TSAFS set to "Disable Cluster: yes" in /ETC/SMS/TSA.cfg as well. The 
other thing we do is copy all the client files to the cluster volumes and run 
the clients from the volumes. We just rename dsmc.nlm to dsmc-vol1.nlm as 
appropriate for the volume in question. This lets us tie the loads and unloads 
of the CADs to the cluster object load/unload script.
  
  If you're dead set on using "dsmc sched" then perhaps there's some specific 
advice on that in the list archives. Oh and of course there's more generic 
advice like keeping current on TSAFS patches.

Andrew Ferris
Network Support Analyst
iCAPTURE Research Centre
University of British Columbia

>>> Peter Schrijvers <Peter.Schrijvers AT BASF-IT-SERVICES DOT COM> 11/9/2006 
>>> 9:03 AM >>>
Hi,

does someone know where I can find good information on backups for novell
cluster volumes.

We have three clusternodes with each 4 volumes.  How can I configure this
backup (prompted mode) so each time a volume is taken over by another
clusternode, the backup schedule gets inactive on the 'old' clusternode'
and get active on the new node?

I'm also having problems unloading the dsmc sched on the volumes.  Server
goes into abend state.  Also an extra push on the ENTER buttom is needed.
So when I want to do this via scripting or something else I may not have to
ENTER manually.

Thx,
Peter Schrijvers




====================
  BASF-IT-Services N.V.
      Peter Schrijvers
          BIS IN-PA
  Tel. +32-3/561.24.77
  Fax +32-3/561.32.20
===================

<Prev in Thread] Current Thread [Next in Thread>