Howard Martin wrote:
Capacity of a compressed tape is dependant on the data being backed up and
the compression algorithm (which can be different on different drives),
with LTO-1 we typically get about 2:1 compression on fileservers and 4:1 on
database type data, with one database I backed up 600GB on one 100G tape
when the database had just been set up it later fell to the 400G region.
Conversly on a standalone DLT8000 I have seen it report compression of
0.9:1!
You are not doing anything wrong the drive manufacturers give the native
capacity and a compressed capacity but they can only assume that any
particular back up will achieve this.
Well yes, but I wouldn't call the case of databases, especially if
they're backed up by modules specific to the database, a case of
compression in the sense that people usually mean it. Databases have
much in common with spare matrices where they can be largely empty
space. And we see that freqently a database dump is way, way smaller
than the database. So the real issue was, I thought, what people
could expect under normal circumstances, which would certainly not be
anything like 260 gb on a 100 gb tape.
On Mon, 24 Nov 2003 10:40:39 -0500, Matt Temple
<mht AT RESEARCH.DFCI.HARVARD DOT EDU> wrote
We've been backing up with compression in various forms for a decade.
I can tell you from our experience that 50 % - 60 % has been most
typical of our compression ratio. I don't think I actually said that
clearly. I think you should expect to get 150 to 160 gb on an AIT-3
tape. On rare occasions, 170 GB.. I've NEVER seen any real-world
compression that looked like
those AIT-3 claims. I'd be most interested in knowing if anyone has.
Then I'd know that we're doing something wrong.
Our data is largely binary image data from microscopes, and scanning
devices of various kinds. Our network is gigabit backboned with
a number of gigabit-capable computers. Our Networker server generally
claims to be running at about 11-12K KB per Sec per drive on a good day
which is fast enough to keep the drives moving, isnt it?
--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
--
=============================================================
Matthew Temple Tel: 617/632-2597
Director, Research Computing Fax: 617/632-4012
Dana-Farber Cancer Institute mht AT research.dfci.harvard DOT edu
44 Binney Street, JF 314 http://research.dfci.harvard.edu
Boston, MA 02115 Choice is the Choice!
--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
