Veritas-bu

[Veritas-bu] NFS Backups

2007-03-27 12:41:27
Subject: [Veritas-bu] NFS Backups
From: David_Cornely at intuit.com (Cornely, David)
Date: Tue, 27 Mar 2007 09:41:27 -0700
So does this Storagetek appliance do snapshots?
I'm not familiar with this piece of equipment but if it's similar to a
NetApp appliance then you might want to consider both disk and tape
backups.

Disk being the built-in snapshot capabilities of the appliance.  Do
something like snapshots during the week and go to tape on the weekend
(of course, you'll have to make sure you've purchased enough disk to
contain all the snapshots for the duration you need to).  If you do this
you might give yourself more time to go to tape so then the NFS option
becomes much easier to implement.

Looking at a combination of things to do will give you more flexibility
towards satisfying your requirements.  I would recommend reducing tape
backups as much as possible though, just less headaches.

-Dave

-----Original Message-----
From: veritas-bu-bounces at mailman.eng.auburn.edu
[mailto:veritas-bu-bounces at mailman.eng.auburn.edu] On Behalf Of Martin,
Jonathan (Contractor)
Sent: Tuesday, March 27, 2007 09:25
To: veritas-bu at mailman.eng.auburn.edu
Subject: [Veritas-bu] NFS Backups


All,

A group I support here is looking very heavily at a Sun Storagetek
appliance like storage device that I've got to backup somehow.  After
seeing the NDMP price tag (we don't do any other NDMP here) the question
of NFS file backups has come up.  So I'm going to try to benchmark NDMP
backups with the NBU NDMP extension (temporary licenses) versus the same
backups on a client server over NFS.

OBVIOUSLY NDMP to Media Server is the way to go here (not an argument
from anyone on that.)  HOWEVER - apparently from a budget perspective
it's a difficult pill to swallow.

(Ok, my disclaimer is complete - so flame on!)

My current plan is to benchmark to NDMP, then load up a test Linux
server and optimize that NFS backup.  I'm assuming the NDMP will be (for
example) some 50+MB/sec and the wire-2-wire is going to be significantly
slower (the percentage slower being directly related to whether or not
this is even feasible.)

Has anyone done any testing like this before?  I'd appreciate any
results anyone can share.

Thanks,

-J

_______________________________________________
Veritas-bu maillist  -  Veritas-bu at mailman.eng.auburn.edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu


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