Re: [ADSM-L] Why virtual volumes?

2007-08-22 13:45:16
Subject: Re: [ADSM-L] Why virtual volumes?
From: Richard Rhodes <rrhodes AT FIRSTENERGYCORP DOT COM>
Date: Wed, 22 Aug 2007 13:47:14 -0400
> I am not understanding the crucial advantage(s) of using virtual
> volumes to backup a data center to a remote site.  Why not backup
> nodes in a remote data center to a TSM server in a local data center?

Sure, do it, we backup many remote sites to our central datacenter.
The question is, what is your DR strategy for the central datacenter
where the TSM server is located?  If that datacenter is destroyed,
where is your backup copy of your data (copy pool) and the
backup of the TSM DB?  The remote nodes are save/secure, but they
are the least of your worry.

Your TSm server needs offsite db backups AND a offsite
copy of your backups (copy pool).

To do electronic offsite copies and db backups, you can
use VV, or implement direct SAN attached tape drives
between our data centers.  VV would use your IP network.  Accessing
remote tape drive directly could be either by extending a SAN
directly or bridging the SAN across the IP network.

Of couse, you don't need to do this with either VV or extended
SAN.  You can also ship copy pool tapes and DB backps offsite
to another facility  for safe keeping.

> We have two data centers about 60 miles apart. If we backup nodes in
> one data center directly to a TSM server in the other the TSM
> architecture is simpler, cheaper and familiar. If we lost the remote
> data center, we could restore its TSM nodes to hardware in the local
> data center thanks to the TSM server already there since it contains
> all the needed metadata. No TSM server build or database restore is
> required. It may be worth mentioning that we are already backing up a
> few remote nodes to a local TSM server with communication speeds in
> the range of other, local backups.

Again, the question is - what happens if you loose that TSM server?

> If we build the virtual volume architecture on the other hand, we
> must purchase, build, and maintain at a distance a second server in
> each data center to connect the tape library there to the TSM server
> in the other data center. If we lose a data center under the virtual
> volume plan, its TSM server and primary disk pool will be lost along
> with everything else. All our eggs will be in one basket. We would
> need to restore the lost TSM server in the other data center before
> we could restore anything else.

True, but how are you going to rebuild the centralized TSM server if
that datacenter is lost?  In a DR you have to rebuild something!

- If you loose the remote node, then you have to rebuild and
restore the node.
- If you loose the central TSM server, you need to rebuild and
restore the TSm server to access your backups.  If this is
a big disaster where you lost the TSm server, library, and all
onsite tapes, then somewhere you need a copy of the backups
(copy pool) and backup of the db, and a location/library/server
to restore to.


The information contained in this message is intended only for the
personal and confidential use of the recipient(s) named above. If
the reader of this message is not the intended recipient or an
agent responsible for delivering it to the intended recipient, you
are hereby notified that you have received this document in error
and that any review, dissemination, distribution, or copying of
this message is strictly prohibited. If you have received this
communication in error, please notify us immediately, and delete
the original message.