ADSM-L

co-location question

2006-09-21 13:14:02
Subject: co-location question
From: John Bremer <jbremer AT LANL DOT GOV>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Thu, 21 Sep 2006 11:13:08 -0600
Greetings,

We are backing up several NetApp NAS filers using TSM NDMP
capabilities.  We have about 700 TB backup data.  Aside from issues
with TOCs, copy pools and some others (which I am confident TSM
development is working to remedy !), we're pretty happy with the service.

Since most NAS volumes are quite large, some multi-terabyte, we have
been co-locating by filespace.  I wonder if anyone out there is
co-locating their NDMP backups by node ?

The reason I want to do this is to break up some filer volumes into
(many, like 100's) virtualfsmapping(s).  This would require a tape
for every virtualfsmapping, and I would like to minimize the number
of tapes per node, but do not want to scatter differentials over too
many tapes, which I suspect will be the case, if I co-locate by
node.  We're using STK 9940B drives, and getting around 300GB per cartridge.

Thanks for any comments.

Regards, John Bremer

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