ADSM-L

Re: Advice needed for EMC CX500 setup

2005-06-07 09:51:09
Subject: Re: Advice needed for EMC CX500 setup
From: PAC Brion Arnaud <Arnaud.Brion AT PANALPINA DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Tue, 7 Jun 2005 15:50:36 +0200
Richard (Aaron too),

Many thanks for taking some of your time trying to help me !

Yes, you're absolutely right, my mail was lacking some important
information, but I did not really knew where to begin ...
To answer some of your questions :

- number of nodes : approx 200, equally split in Windows and AIX
environments. Some of them also hosting Informix and DB2 databases. 
- backups/day : approx 400 GB AIX/Win data, and 500 to 750 GB
informix/db2. No archiving (except db2 logical logs), no hsm.
- db actual size : 30GB, 80% used; expected 50% increase within one year
(proportional to the expected increase of nodes). Number of objects
(from expiration process) : 6100000. Our actual performance for
expiration is approx 1600000 obj/hr, not much, but due to time
constraints space reclamation is running parallely and slows it. I would
expect getting the same performance, or better !
- log size : 13 GB for security as I'm using rollforward mode
- expected iops : well, no idea at all. Some formula to calculate this ?

Do you need something else ?
Regards.

Arnaud 

************************************************************************
******
Panalpina Management Ltd., Basle, Switzerland, CIT Department
Viadukstrasse 42, P.O. Box 4002 Basel/CH
Phone:  +41 (61) 226 11 11, FAX: +41 (61) 226 17 01
Direct: +41 (61) 226 19 78
e-mail: arnaud.brion AT panalpina DOT com
************************************************************************
******

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Richard Rhodes
Sent: Tuesday, 07 June, 2005 14:46
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Advice needed for EMC CX500 setup

Hi Brion,

I see that no one else has responded . . . . . let me stir up the waters
and see what shakes loose . . .

You don't inidicate the the number of nodes, db objects (files), backup
rate (gb/day), etc, etc that you expect.  This is key.  You indicate a
db of 70 gb with a full sized log - this means to me that you expect
heavy db activity.  For the tsm db, or for any other db for that matter,
the question is how many iops do you need?  The 133gb fc disk drive can
probably give you around 120+ iops.  I'm afraid that a 3 spindle raid 5
probably is not enough iops to run tsm instance with the activity you
might have.

Before my time on the storage team, Emc sold a disk subsystem to us for
one of our TSM servers.  It was a Clariion with all big drives.  Like
most disk purchases, it was configured on capacity, not i/o load.  There
was no way to split the drives between db/log and staging pools and have
enough disk space for staging pools, or, enough spindles for db/log.
The only way to make it work was to spread db/log/pools across all the
spindles.  NOT good . . . . actually, it worked quite well for the first
year.

Rick








                      PAC Brion Arnaud
                      <Arnaud.Brion@PAN        To:
ADSM-L AT VM.MARIST DOT EDU
                      ALPINA.COM>              cc:
                      Sent by: "ADSM:          Subject:  Advice needed
for EMC CX500 setup
                      Dist Stor
                      Manager"
                      <[email protected]
                      .EDU>


                      06/06/2005 01:13
                      PM
                      Please respond to
                      "ADSM: Dist Stor
                      Manager"






Hi list,

I need some guidance for setup of a brand new EMC array, equipped with
10 FC drives, 133GB each. TSM server (5.3.1) will be running on AIX 5.3.
My goal is to use those disks for following purposes :
- TSM DB 70 GB
- TSM log 13 GB
- TSM primary pools : all of the remaining space

Questions :

1) EMC recommands using RAID3 for disk to disk backup -> should I create
2 raid groups : one raid5 for TSM DB and logs, one raid3 for diskpools ?

2) if using raid5 for DB and log, should I create 2 raid groups, each
one using different physical disks : one for the DB, the other for the
logs, or is the performance penalty neglectible if both are on the same
disks ?
3) I plan using raw volumes for DB and LOGS (no TSM or AIX mirroring, I
trust EMC's raid): what is the best setup : 1 big TSM volume or several
smaller ?  If using several volumes should they be built using different
LUN's, or could I create several LV's on the same LUN.

So many possibilities, that I'm a bit lost ...  So if you have good
experience in this domain, your help will be welcome !
TIA.
Cheers


Arnaud

************************************************************************
******
Panalpina Management Ltd., Basle, Switzerland, CIT Department
Viadukstrasse 42, P.O. Box 4002 Basel/CH
Phone:  +41 (61) 226 11 11, FAX: +41 (61) 226 17 01
Direct: +41 (61) 226 19 78
e-mail: arnaud.brion AT panalpina DOT com
************************************************************************
******




-----------------------------------------
The information contained in this message is intended only for the
personal and confidential use of the recipient(s) named above. If the
reader of this message is not the intended recipient or an agent
responsible for delivering it to the intended recipient, you are hereby
notified that you have received this document in error and that any
review, dissemination, distribution, or copying of this message is
strictly prohibited. If you have received this communication in error,
please notify us immediately, and delete the original message.

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