Veritas-bu

[Veritas-bu] Fewer pairs = better dup performance?

2003-11-04 09:12:32
Subject: [Veritas-bu] Fewer pairs = better dup performance?
From: Mark.Donaldson AT experianems DOT com (Donaldson, Mark)
Date: Tue, 4 Nov 2003 07:12:32 -0700
This is one of those maybe/maybe-not questions.  The answer all hinges on
whether your network link is your bottleneck.  Buffering, client speeds,
incrementals vs. fulls, etc. all play into it, too.

Your best option, IMO, is just to try it.

We here, by the way, have driven FC speeds to better than 75 MB/sec.
(Blocking large will increase FC transfer rates).

-M

-----Original Message-----
From: Shafto, Eric [mailto:Eric.Shafto AT drkw DOT com]
Sent: Tuesday, November 04, 2003 6:32 AM
To: veritas-bu AT mailman.eng.auburn DOT edu
Subject: [Veritas-bu] Fewer pairs = better dup performance?


I'm using a Solaris master with a single Gbit HBA to talk to my tape drives.
I have read here (and it seems to be true) that there is a maximum aggregate
throughput of about 60MBps. Since I get a little better than 15MBps to each
drive, two pair of drives just about consumes all my throughput.

Right now I'm using three pair, and they are averaging slightly better than
10MBps. My question is, would I get better throughput if I cut down to two
pair, and (presumably) reducing the amount of shoe-shining?

Any thoughts on this will be appreciated.


If you have received this e-mail in error or wish to read our e-mail
disclaimer statement and monitoring policy, please refer to
http://www.drkw.com/disc/email/ or contact the sender.

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

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