Hi Andy

I need to set up diskpools for windows directories and SQL Meta data.  I
have an idea that lots of small entries will be better on diskpool than
sequential files.  Also sequential files can't reuse space until
reclamation is run, which wastes space and uses a lot of machine
resources for the reclamation jobs.

I figure to use diskpools for data that will live and die on disk and
never make it to tape.


Regards

Steve.

Andrew Carlson wrote:

Could you explain why you feel the need to use disk pools vs
sequential disk files?  We are going the route of sequential disk file
storage pools, and I haven't run across anything that would make me
want to keep disk pools for certain uses.  Thanks a lot.

On Thu, Apr 22, 2010 at 3:48 AM, Steven Harris <steve AT stevenharris DOT info> 
wrote:

Hello All,

Conventional wisdom is that for TSM on Solaris, raw volumes are the way
to go.  I have 4 Sun X4540 servers each with 64GB of memory and 48x1TB
internal SATA disks.  The whole disk farm is one big ZFS zpool.  TSM DB
and Log are on external fast SAN disk.

So far I've been using sequential file pools, but feel the need for some
disk pools for windows directories and MSSQL Metadata.  Is ZFS still as
inefficient as UFS was for this purpose?

Regards

Steve.

Steven Harris
TSM Admin, Paraparaumu, NZ

------------------------------------------------------------------------


No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 9.0.814 / Virus Database: 271.1.1/2829 - Release Date: 04/23/10 
06:31:00