ADSM-L

Re: TSM DB/Disk Pool - Disk tuning question

2001-08-30 12:52:58
Subject: Re: TSM DB/Disk Pool - Disk tuning question
From: Tab Trepagnier <Tab.Trepagnier AT LAITRAM DOT COM>
Date: Thu, 30 Aug 2001 11:53:29 -0500
Lindsay,

The "locked volume" is my understanding based on what I've observed and
picked up from the list and the manuals. If that is wrong, I apologize.

My set of 500 MB volumes is spread across six physical disks so I do get
some load balancing.

I think my system is network-bound anyway. The TSM server is on a 100 Mbs
link.  Running topas, I've seen the newest IBM 18 GB SCSI drives WRITE at
12 MB/s.  Since almost all of my disk pool volumes are on 18Gs, I doubt
that disk performace will limit the system's throughput.

Thanks.

Tab







Lindsay Morris <lmorris AT SERVERGRAPH DOT COM>@VM.MARIST.EDU> on 08/30/2001
11:42:57 AM

Please respond to lmorris AT servergraph DOT com

Sent by:  "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>


To:   ADSM-L AT VM.MARIST DOT EDU
cc:
Subject:  Re: TSM DB/Disk Pool - Disk tuning question


Do sessions really lock volumes?

I had always assumed they didn't.  Disk volumes are random access, so for
some reason I thought that several sessions could share one disk volume.
In
fact, I know of one site that regularly has 45-50 backup sessions running
simultaneously, but only 37 disk volumes... so I'm not sure that's true.

I DO believe that each new session tries to get a NEW disk volume to use;
thus TSM tends to balance out the nightly backup load across its disk
volumes.  Maybe that's the behavior you were seeing?

If that's right (Andy Raibeck? is it?), then putting 36 500MB volumes on
one
18 GB disk would encourage TSM to use that one disk for all of its backup
sessions - a bad thing.

The more common wisdom is to make volumes that take up the entire physical
disk.  The only problem there is that you have less flexibility: if, in the
future, you need to reapportion space between, say, your DB and a storage
pool, you have to move  an entire volume.  I like to cut things up into 2
GB
chunks; that avoids the requirement for a BFS-enabled filesystem, too.


> -----Original Message-----
> From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU]On Behalf Of
> Tab Trepagnier
> Sent: Thursday, August 30, 2001 12:26 PM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: TSM DB/Disk Pool - Disk tuning question
>
>
> David,
>
> My preference for disk pools is to use a lot of smaller volumes. In our
> case we use 500 MB volumes in whatever quantity is required to
> get the disk
> pool to its desired size.  The reason is that sessions lock volumes.  A
> small number of disk volumes works the same as a small number of tape
> drives.  More volumes means more simultaneous sessions.
>
> As for database volumes, my database (~21 GB) is spread across four pairs
> of mirrored disks.  You can use Q DBV F=D to see the percent
> utilization of
> each DB volume.  If they are unbalanced, you can use *SM's feature of
> copying DB pages elsewhere when deleting DB volumes to force balance the
> volume utilization.  I've done that and I have the DB volume utilization
> almost perfectly balanced between the four pairs of disks.
>
> Good luck.
>
> Tab Trepagnier
> TSM Administrator
> Laitram Corporation
>
>
>
>
>
>
>
> David Longo <David.Longo AT HEALTH-FIRST DOT ORG>@VM.MARIST.EDU> on 08/30/2001
> 09:27:04 AM
>
> Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
>
> Sent by:  "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
>
>
> To:   ADSM-L AT VM.MARIST DOT EDU
> cc:
> Subject:  TSM DB/Disk Pool - Disk tuning question
>
>
> I have TSM Server 3.7.4.0 on AIX 4.3.3.  I am going to be redoing my disk
> storage for this machine and have basically two questions. if someone
> knows the details of how this works.
>
> 1.  The Disk Storage pool I have is 110GB.  Most of the time it is
between
> 20 and 70 percent full - mostly less than 50 percent.  If I have
> this split
> into say 10 volumes, each on a separate hard disk, does most of the
> activity stay on the lower volumes and rarely on the higher ones?
>
> Logically I see this as a coffee cup with a slow leak.  The bottom is
> almost always wet, the top rarely.  Is my assumption correct?
>
> Basically I would spread the volumes over the hard disks to even out disk
> access.
>
> 2.  Same basic question about TSM DB.  It is 13GB and 65% utilized.  I
> imagine like a regular database it is spread out over the 13GB and has
> more or less even access if spread out over several disks.  Correct?
>
> (With my current system the way everything is laid out now I can't use
> things like "iostat" to prove/disprove this, and I didn't seem to
> find this
> "specific" topic mentioned in manuals.)
>
> Thanks,
>
>
> David B. Longo
> System Administrator
> Health First, Inc.
> 3300 Fiske Blvd.
> Rockledge, FL 32955-4305
> PH      321.434.5536
> Pager  321.634.8230
> Fax:    321.434.5525
> david.longo AT health-first DOT org
>
>
>
> "MMS <health-first.org>" made the following
>  annotations on 08/30/01 10:33:23
> ------------------------------------------------------------------
> ------------
>
> This message is for the named person's use only.  It may contain
> confidential, proprietary, or legally privileged information.  No
> confidentiality or privilege is waived or lost by any mistransmission.
If
> you receive this message in error, please immediately delete it and all
> copies of it from your system, destroy any hard copies of it, and notify
> the sender.  You must not, directly or indirectly, use, disclose,
> distribute, print, or copy any part of this message if you are not the
> intended recipient.  Health First reserves the right to monitor all
e-mail
> communications through its networks.  Any views or opinions expressed in
> this message are solely those of the individual sender, except (1) where
> the message states such views or opinions are on behalf of a particular
> entity;  and (2) the sender is authorized by the entity to give such
views
> or opinions.
>
> ==================================================================
> ============
>