Networker

Re: [Networker] de-duplication

2008-11-05 07:14:28
Subject: Re: [Networker] de-duplication
From: Francis Swasey <Frank.Swasey AT UVM DOT EDU>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Wed, 5 Nov 2008 07:09:50 -0500
On 11/5/08 12:42 AM, tkimball wrote:

It may also be possible to use something like the Cypress (GreenBytes) NAS and 
use its internal de-dup/compress for adv_file devices (over iSCSI).  Networker 
wouldn't know, or care.

http://green-bytes.com/cypress.html

Datadomain also plays in this space -- We have used their equipment for several years now. The one gotcha with dedup is that you must write your own scripts if you want to implement staging. Because of the de-dup process, moving a 1TB saveset off the aftd to another medium will not (or at least shouldn't) free up 1TB of space on the de-dup appliance. I've asked EMC if their Avamar based de-dup product has solved that issue -- no answer yet.
We're trying to get one here in order to test that theory, since it *may* be 
able to replace the two *racks* of StorageTek Flex disk we use for adv_file 
today (we have a Tier 4 DBO).

Of course, the sales folks got real silent when I mentioned using it in 
environments with pre-compressed data (our Sybase dumps, which is about 20 TB 
currently - about half the 'real' disk space of the Thumper chassis that 
Cypress uses).  I wanted to know how their de-dup extensions to ZFS would work 
out here, since it was a proprietary change to the code.


Yeah, they should get quiet. Pre-compressed data should look different every day -- there is very little chance (practical experience here with datadomain devices and pre-compressed data) for pre-compressed data to be de-dup'd or even compressed further at all. In general, compressing compressed data makes it bigger.

--
Frank Swasey                    | http://www.uvm.edu/~fcs
Sr Systems Administrator        | Always remember: You are UNIQUE,
University of Vermont           |    just like everyone else.
 "I am not young enough to know everything." - Oscar Wilde (1854-1900)

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

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