ADSM-L

Re: ? Should we set Resourceutilization > 10 if appropriate ?

2006-06-05 18:42:00
Subject: Re: ? Should we set Resourceutilization > 10 if appropriate ?
From: Remco Post <r.post AT SARA DOT NL>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Tue, 6 Jun 2006 00:41:37 +0200
James R Owen wrote:

> Management discourages use of undocumented/unsupported settings,
> but I'm arguing that we need to specify RESOURCEUTILIZATION 30
> in order to effect efficient backups for our email servers:
>     4 IMAP servers, each w/4 CPUs, running linux client 5.2.3,
>     each backs up 15 FS sending ~200GB/night (compressed)
>     via 100Mb -> Gb ethernet
>     to our TSM 5.2.3 service's disk stgpool
>

I guess your problem is that finding a small amount of changed files in
these large filesystems. I solved this by:

1- generating a list of virtualmountpoints, basically, TSM is still
faster in doing an incremental of 100 filesystems of 100,000 files each
than (even mutithreaded) 1 filesystem of 10,000,000 files.

2- doing an 'incremental bydate' during weekdays and only normal
incrementals during the weekend.

also, you might want to look into the type of filesystem you are using.
In these cases reiserfs could perform better than any other filesystem.

Basically, My guess is that increasing your resourceutil is not that
usefull. in these cases the disks just don't keep up, whatever you do.

If you do want to control the order things are done in, my bet would be
on the domain statement, of course, you need to update that in case you
add a filesystem (I've heard of scripts that do this).

--
Met vriendelijke groeten,

Remco Post

SARA - Reken- en Netwerkdiensten                      http://www.sara.nl
High Performance Computing  Tel. +31 20 592 3000    Fax. +31 20 668 3167
PGP Key fingerprint = 6367 DFE9 5CBC 0737 7D16  B3F6 048A 02BF DC93 94EC

"I really didn't foresee the Internet. But then, neither did the
computer industry. Not that that tells us very much of course - the
computer industry didn't even foresee that the century was going to
end." -- Douglas Adams

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