Re: Multiple SQL version count with cusror_sharing=similar

From: Neeraj Bhatia <neeraj.dba_at_gmail.com>
Date: Wed, 1 Jul 2009 10:26:42 +0530
Message-ID: <a8fd4d730906302156p6a8fe941x92b5a7dbb6ca1360_at_mail.gmail.com>

Hi,

First of all thanks a lot Cheng for the clue.

I just turned-off dynamic sampling and it behaved as expected. Means, 1 version_count in case of similar and exact cursor_sharing setting.

But, even in case of dynamic sampling, table data is not skewed and histogram is of no use in this case. Thi can be checked even from plan_hash_value of child cursors. Execution plan is same. Then why Oracle is creating multiple child.
Disabling the bind value peeking, resulting in three parent cursors. I didn't know any reason for that.
- Neeraj

--

http://www.freelists.org/webpage/oracle-l Received on Tue Jun 30 2009 - 23:56:42 CDT

This message: [ Message body ]
Next message: Herald ten Dam: "RE: bulk collect into ... limit N"
Previous message: Michael Wehrle: "Re: Quickest Way To Recreate a Database"
Maybe in reply to: Neeraj Bhatia: "Multiple SQL version count with cusror_sharing=similar"

Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [ by author ]

Original text of this message