Networker

Re: [Networker] Query in Staging from adv_file

2008-08-21 15:00:06
Subject: Re: [Networker] Query in Staging from adv_file
From: Stan Horwitz <stan AT TEMPLE DOT EDU>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Thu, 21 Aug 2008 14:55:53 -0400
On Aug 21, 2008, at 2:39 PM, Curtis Preston wrote:

Francis Swasey said:

As currently implemented, you are correct.  A stage or a clone or
multiple recovers may be performed on the AFTD. If a stage is running,

clones and recovers have to wait for it to complete.

Hmmm... VTLs don't have that problem. ;)  OK, they sort of do.  You
can't restore from a tape that's currently being cloned, but you can
restore from other tapes.  It looks like the entire AFTD is held up
while cloning/staging is going on.

Assuming you were using 10 GB virtual tapes (and I think you should),
this would cause a maximum of a 3 minute delay for a given restore if
you were using a VTL and you had tape contention. Here's my math: A 10
GB * 2:1 compression = a 20 GB tape.  At a clone speed of 100 MB/s
(which you should easily get if you bought the right one), that tape
will be cloned/destaged in about 3 minutes, allowing it to be used for a
waiting restore.

Another AFTD limitation that I've heard of, but haven't verified is that
the RO side of the AFTD is counted in some way against your device
count.  I don't know the details and haven't verified, but there are
some per server device limits that certain versions of NW have, and my
understanding is that the AFTD is counted as two against those numbers.
Can anyone verify that this is true or false?

I don't think device count is relevant in a NetWorker environment because EMC doesn't license anything based on device count. What I am starting to see is that a VTL would be better for my environment because it would be easier to carve out the storage space into lots of little (10GB VTs) instead of having larger ADFTs. I figure that with the law of large numbers on my side, the odds of needing to recover data from lots of virtual tapes with them all staging is much slimmer then my needing to recover data from far fewer, but larger ADFS devices that are all busy staging data when I need to recover something.

To sign off this list, send email to listserv AT listserv.temple DOT edu and type 
"signoff networker" in the body of the email. Please write to networker-request 
AT listserv.temple DOT edu if you have any problems with this list. You can access the 
archives at http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER