RE: Materialized View+complete refresh

From: Powell, Mark D <mark.powell_at_eds.com>
Date: Thu, 28 May 2009 15:17:45 -0400
Message-ID: <D1DC33E67722D54A93F05F702C99E2A903DDCD18_at_usahm208.amer.corp.eds.com>


 

Just because you have tuned the query does not mean that more gains cannot be had by re-tuning it again.

If the query is as good as it can get on its own then you have adding parallelism but to get the best gains from running parallel you might actually have to un-tune the query. That is, some of the changes you made to get the query to run in 20 minutes may actually hinder the CBO with PQO so you may need to test several versions of the query.

Otherwise you need to try to tune the refresh process as Jared suggested.

But where is the explain plan for this query? Everyone is working blind without it.

  • Mark D Powell -- Phone (313) 592-5148

-----Original Message-----

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Mohammed Mehraj hussain
Sent: Wednesday, May 20, 2009 10:29 PM
To: Jack van Zanen
Cc: oracle-l_at_freelists.org
Subject: Re: Materialized View+complete refresh

we Have already tuned the query and then we have reduced it to this time(20 min)...
--

http://www.freelists.org/webpage/oracle-l

--

http://www.freelists.org/webpage/oracle-l Received on Thu May 28 2009 - 14:17:45 CDT

Original text of this message