Re: Huge tempspace requirement - doesn't match explain plan
From: Greg Rahn <greg_at_structureddata.org>
Date: Fri, 22 Aug 2008 03:19:37 -0700
Message-ID: <a9c093440808220319r44ace51do3f9b434484ee8015@mail.gmail.com>
Date: Fri, 22 Aug 2008 03:19:37 -0700
Message-ID: <a9c093440808220319r44ace51do3f9b434484ee8015@mail.gmail.com>
First things first: are the cardinality estimates from the plan
accurate with the actual number of rows from each row source?
If they are not, it would easily explain why the temp space estimate
is much more than the actual.
I suggest you break it down into bite-size chunks. Start with the filters on each table to check the filter cardinality. Then start with a two table join, then three, then four, etc. monitoring the join cardinality with each to validate its accuracy.
My notes on plan troubleshooting:
http://structureddata.org/2007/11/21/troubleshooting-bad-execution-plans/
--
Regards,
Greg Rahn
http://structureddata.org
--
http://www.freelists.org/webpage/oracle-l
Received on Fri Aug 22 2008 - 05:19:37 CDT