Hello,
there are some problems with expiration at 3.1.2.40
level of the server. The first one, IY05372 addresses
an intermittent hang during expiration. This apar is fixed
at the 3.1.2.42 level of the server. Another apar of
interest would be IY05562, deadlock during expiration.
The fix for this apar is not yet available but should be
soon. You should consider going to 3.1.2.42.
That should help your expiration process performance.
Since you mentionned you saw directories after
directories from the archives table then you may have
duplicate directories. The duplication of archive
directories was a problem introduced at v3. This issue
is discussed in apar IX89638, which is fixed at the 3.1.2.40
level of the server. You can remove duplicate archive
directories using the
CLEAN ARCHDIRectories ...
command. Please refer to the 3.1.2.40 README file
for discussion on this new adsm admin command.
You also need your client to be at the 3.1.0.7 to stop
duplicate directories from being created again.
That should also help your expiration process performance.
Finally, please take a look at your database cache hit % ratio.
You may need to increase the database buffer pool size.
A bigger BUFPOOLSIZE will require more memory but will
also help expiration performance.
Good day !
-----------------------------------------------------------------
Rejean Larivee
Rejean Larivee
IBM ADSM Level 2 Support
Arshad Munir <arshadmunir AT HOTMAIL DOT COM> on 11/28/99 02:55:47 PM
Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
cc:
Subject: Archive Directories & Expire performance
Hello ADSMers,
I have yet another wonderful experience, I had been watching
my expiration process for a while, it started with few minutes
now it is few days !!! yes believe it or not.
I have a 180 NT client hooked up to an NSM with Server-to-Server
to HOST which has VTS. I have been doing some tests as when I got
a record expiration time to one node of 32hours !!!!. I started
filtering what is going on & suddenly I realized that the when
expire process goes to Standard Mgmt Class for me which is
Default, it stays there for hours. Than as next step I did select
to see what do I have in there & saw directories & directories
from archives, some of them with 26 versions & few of them
appreared 4 times with same name, date & time.
So as second step just to get the expiration process going
I increased the retension period of the class & my expiration
goes fine.
I have clients starting from 3.1.0.0 to 3.1.0.7 my server is
3.1.2.40 , has someone has experienced this before.
I have following doubts,
1. How do I do clean up?
2. Do my database has fragmentation?
3. DO I have to reorgnaize my table? some DB2 person told me that
one has to do this in order to improve performance.
I cannot relay on expire process as my robot is filling up
much faster than my expiration process.
Can anyone help me in this regard,
thanks in advance
Arshad Munir
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
|
