ADSM-L

Re: [ADSM-L] DBUnload, need for speed

2012-06-27 10:05:01
Subject: Re: [ADSM-L] DBUnload, need for speed
From: Richard Rhodes <rrhodes AT FIRSTENERGYCORP DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Wed, 27 Jun 2012 09:59:48 -0400
> The first part of the unload is running very fast (4 or even 5 million
> entries per minute, but at a given point performance drops dramatically.
> It's always at the same point (around 80%), so my guess is that that's
> the most heavily fragmented part.
> The only thing I can think of to speed up the process is to use a
> storage device with a very low latency, like SSD, but unfortunately
> that's not supported on the P5 series...

So, what is slowing down?

I'll throw out some questions:
- During the slowdown . . .
    Is AIX waiting on disk I/O?  (I/O wait state)
    Is AIX CPU running hard?     (some cpu intensive task)
    What does iostat see as response time?
      (iostat -DlTR <interval> <count>)
- VMAX
    Check VMAX SPA for the time during the run, or watch real time during
a test run.
      You can see the r/w iops and r/w resp times for the storage group of
you TSM db/log.
    Do you have a SDD tier in you VMAX?  Is there enough room to move
      some/all of the TSM db/log up into it?
    If you are running FASTVP with a sata tier, move the db/log off
      of the SATA tier.
    Are the FA's and/or FA ports overloaded?

Rick


-----------------------------------------
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>