RE: 10g RAC and defined services performance problem
Date: Wed, 22 Oct 2008 18:04:23 -0500
Message-Id: <20081022230434.17E9FA91217@turing.freelists.org>
[ problems with services moving back and forth spontaneously ]
Are you saying that the clusterware is failing the services back and forth
between nodes spontaneously? If that is happening frequently it would
explain the performance problems. You might want to look in the cluster logs
for anything suspicious. Perhaps some problems with the interconnect that
are causing the cluster services to fail connections. It might also be worth
while to trace a session during the problem period and see what is going on
at that level.
-Randy
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
On Behalf Of Crisler, Jon
Sent: Friday, September 12, 2008 11:24 AM
To: adar666_at_inter.net.il
Cc: oracle-l
Subject: RE: 10g RAC and defined services performance problem
Dan Norris and I have been talking about this, and the most obvious performance issue is really with non-sharable SQL code. So far it appears that sessions are staying on the designated node, but we do have problems with services moving back and forth spontaneously. Unfortunately I don't have enough info at this point to do further diagnosis but I should have that in 1-2 days.
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
On Behalf Of Yechiel Adar
Sent: Friday, September 12, 2008 2:33 AM
Cc: oracle-l
Subject: Re: 10g RAC and defined services performance problem
I think that the first thing to do in this configuration is: select * from gv_$session to be sure that what I planned is working and sessions are not moved around in the RAC.
The second thing is to get AWR report for two or three sample hours and see the waits and logical gets to find waiting and/or heavy consumers.
I would also get online during working hours and check for locks from time to time.
Adar Yechiel
Rechovot, Israel
Crisler, Jon wrote:
We have a 3-node 10.2.0.3 RAC Server on Red Hat 4, and have defined about 20 additional services with preferred nodes set to 1 specific node, with available nodes set for the other 2 nodes. Workload is divided evenly with about 7 services on each node. Has anybody run into specific problems with performance when using a lot of services, as opposed to having the application just connect to the main db load-balanced service ?
This particular application has a reputation from the developers of not scaling well across RAC nodes, and we have recently been advised of a lot of performance problems that I am now starting to investigate, but I want to get some opinions on the wisdom of using many different services rather than just a few or one.
-- http://www.freelists.org/webpage/oracle-lReceived on Wed Oct 22 2008 - 18:04:23 CDT