ADSM-L
[Top] [All Lists]
<prev [Date] next>
[Advanced]
 <prev [Thread] next>

Re: [ADSM-L] 3592 drives no compression

2009-10-26 10:27:34
Subject: Re: [ADSM-L] 3592 drives no compression
From: Richard Sims <rbs AT BU DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Mon, 26 Oct 2009 10:26:09 -0400 
On Oct 26, 2009, at 9:42 AM, Lars-Erik Öhman wrote:
It´s generation 1 drives, and the driver is the latest from IBM (September 2009).
I check in device manager and the driver version is 6.2.0.1 (3592J1A), and it even say in "Tape drive parameters" that Compression capable is TRUE, and Compression enabled TRUE. 

Okay - that helps narrow it down.
You indicate that the client options do not specify that the client should be doing compression. Do some 'dsmc query options' to assure that. Another thing to check is whether the REGister Node that was used for these clients specified COMPression=Yes - or check the session summary report from recent backups, for compression indications.
Beyond that, the cause is obscure. I would pursue by defining a temporary storage pool and mgmt class (for tape isolation) performing a backup on a very large, ordinary-content file, look at the backup statistics, then perform a Select on it in the Backups table to look for comparable size there, to see if there is any compression in the path to TSM. Outside of TSM, you might take a drive from TSM, mount a spare 3592 tape via your OS and have a program write ordinary data to it until EOT was reached, where it would report its progress, and you could then assess compression. 

   Richard Sims
[More with this subject...]
<Prev in Thread] Current Thread [Next in Thread>
Previous by Date:  Re: [ADSM-L] Performance and migration: AIX vs Linux, Allen S. Rout
Next by Date:  Re: [ADSM-L] TSM journaling, Howard Coles
Previous by Thread:  Re: [ADSM-L] 3592 drives no compression, Lars-Erik Öhman
Next by Thread:  Re: [ADSM-L] 3592 drives no compression, Lars-Erik Öhman
Indexes:  [Date] [Thread] [Top] [All Lists]