Amanda-Users

Re: question on Tapetypes

2003-10-01 18:51:44
Subject: Re: question on Tapetypes
From: Jon LaBadie <jon AT jgcomp DOT com>
To: amanda-users AT amanda DOT org
Date: Wed, 1 Oct 2003 18:45:48 -0400
On Wed, Oct 01, 2003 at 02:54:34PM -0700, Jeanne Case wrote:
> I recently installed a Dell PowerVault 110T 320 SDLT tape drive. I was 
> not able to find a tapetype for this unit. So I ran the amtapetype test. 
> My question is if I plan on using this drive with hardware compression 
> on, should I still run the amtapetype test with the hardware compression 
> off?

amtapetype should always be run with compression off.

In fact one of the features of amtapetype vs the older tapetype
is detection at the start if compression is on.

However, given that you plan to use hardware compression,
any number you get from amtapetype is a lower boundary.
But you already know approximately what it will be, 150-160GB.
And you will certainly want to fudge the number for your
anticipated actual compression.

Since you already know upper and lower boundaries (320GB
and 160GB) and you will have to fudge (guess) at the actual
hardware compression, and whatever amtapetype gives you will
be nearly useless for your situation, pick a number between
the bounds.  About halfway (say 250GB) is always a good
starting point.  Then if you find lots of times it won't
hold 250GB because your data is incompressible (amanda
will tell you how much it sent to the drive) you can adjust.
Or adjust upward until the tapes do start to fill.

The other numbers from amtapetype are of minor or no import.
"speed" is totally unused by amanda.  It is easy to calculate
from the data amtapetype collects and often is of interest to
the admin.  "filemark" today is such a small percentage of the
total tape capacity (which you are guessing at anyway) that
you could set it to zero with no harm.

jl
-- 
Jon H. LaBadie                  jon AT jgcomp DOT com
 JG Computing
 4455 Province Line Road        (609) 252-0159
 Princeton, NJ  08540-4322      (609) 683-7220 (fax)

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