10g Resource Management Plan changes of its own volition

From: Martin Klier <usn_at_usn-it.de>
Date: Tue, 30 Sep 2008 18:07:42 +0200
Message-ID: <48E24ECE.2000502@usn-it.de>

Dear list,

I've got trouble with my resource management. It's a RAC x86_64 on Linux (SuSE) with four nodes.

First, I just noticed that only node 1 used the user defined plan MY_PLAN (just an example) regularly. Looking up the plan config in EM Grid Control, it was only configured for Node 1, Nodes 2-4 always fell back to INTENAL_PLAN which is the default. I changed this several times in EMGC. (Much too late I recognized that it's just a system parameter, shame on me)

So now I did set it this way:
alter system set resource_manager_plan=MY_PLAN scope=both sid='*'; as well as for each sid seperately for a try.

It worked like a charm, too-much-consuming statements suffered from CPU quota wait events now. But only for some time, at once all nodes had set this value for resource_manager_plan: "SCHEDULER[0xC1E6]:INTERNAL_PLAN" And, needless to tell, MY_PLAN was out of use again.

First of all, why is the system changing my parameters? And who, which process does that? The scheduler? Why should it do so?

Thanks if anybody has an idea...
Martin Klier

Usn's IT Blog for Linux, Oracle, Asterisk http://www.usn-it.de


http://www.freelists.org/webpage/oracle-l Received on Tue Sep 30 2008 - 11:07:42 CDT

Original text of this message