ADSM-L

Re: SSA Disk performance

2005-02-02 12:41:37
Subject: Re: SSA Disk performance
From: "Mark D. Rodriguez" <mark AT MDRCONSULT DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Wed, 2 Feb 2005 11:41:01 -0600
Hi,

When using LVM mirroring you need to consider a few things.  First of
all judging by your symptoms I would say that your "Scheduling Policy"
for that LV is not set to parallel.  In parallel mode it will read the
data from the drive that will respond the quickest.  In sequential mode
it always reads from the first copy unless there is a problem with that
read and then it moves to the second copy.  There is a variant of
sequential "parallel write/round robin read" this time as the name
suggests it round robins through the mirrors for the reads.  Another
performance hit can be if you are using the "Mirror Write Consistency
(MWC)".  MWC does not buy you that much from an availability stand point
but does cost in performance.  Furthermore, if you SSA cards have the
battery backed up cache card it really is over kill to have MWC on at
all.  If you VG has many disks in it you should consider doing RAID 10,
that is striped and mirrored, which will give you better performance.
Although I am not sure if you can do that in AIX 4.3.3 I don't quite
remember when they started supporting doing striped and mirroring
through the LVM.

Another comment about raw vs lvm, raw is not always faster then lvm.  In
fact quite often lvm will out perform the raw partition!

Now on to tunning, if you are still not where you need to be there is a
great deal more performance tuning that you can do.  Try this first, if
it is not what you need then if you search the archives there have been
numerous discussion about tuning an AIX OS for a TSM environment.  If
those previous posts are not clear to you repost with your questions and
I will see if I can help.

--
Regards,
Mark D. Rodriguez
President MDR Consulting, Inc.

===============================================================================
MDR Consulting
The very best in Technical Training and Consulting.
IBM Advanced Business Partner
SAIR Linux and GNU Authorized Center for Education
IBM Certified Advanced Technical Expert, CATE
AIX Support and Performance Tuning, RS6000 SP, TSM/ADSM and Linux
Red Hat Certified Engineer, RHCE
===============================================================================



Loon, E.J. van - SPLXM wrote:

Hi *SM-ers!
We are currently struggling with our SSA disk performance.
For some strange reason we see that although all disks use LVM mirroring,
the reads are not balanced. In most case one disk is used for 60 or 70%
while it's mirror is just using 3%.
We also measured the disks performance. We created a file on a raw logical
volume and measured around 15 Mbytes/sec. We created a JFS filesystem on the
same disks and reran the test: 40 Mbytes/sec. I cannot explain this
difference! RAW should perform better that JFS on AIX 4.3.3, right?
Today we upgraded AIX to the latest maintenance level (11) and installed the
latest firmware on the disks, but performance hasn't improved.
I was hoping that TSM-ers with AIX performance knowledge can help me out
here, I'm stuck right now...
Thank you VERY much in advance for any reply!!
Kindest regards,
Eric van Loon
KLM Royal Dutch Airlines


**********************************************************************
For information, services and offers, please visit our web site: 
http://www.klm.com. This e-mail and any attachment may contain confidential and 
privileged material intended for the addressee only. If you are not the 
addressee, you are notified that no part of the e-mail or any attachment may be 
disclosed, copied or distributed, and that any other action related to this 
e-mail or attachment is strictly prohibited, and may be unlawful. If you have 
received this e-mail by error, please notify the sender immediately by return 
e-mail, and delete this message. Koninklijke Luchtvaart Maatschappij NV (KLM), 
its subsidiaries and/or its employees shall not be liable for the incorrect or 
incomplete transmission of this e-mail or any attachments, nor responsible for 
any delay in receipt.
**********************************************************************




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