Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Influence Optimzer to merge 'IN' subquery

Re: Influence Optimzer to merge 'IN' subquery

From: Sybrand Bakker <sybrandb_at_hccnet.nl>
Date: Wed, 09 Feb 2005 19:27:50 +0100
Message-ID: <pblk01lpbml6kpkj4q73at88o7e87reld0@4ax.com>


On 9 Feb 2005 08:29:50 -0800, mccmx_at_hotmail.com (Matt) wrote:

>
>So how can I force Oracle to evaluate the subquery first and use the
>value as part of the index range scan of the large table.

There is PUSHQ or a PUSHD hint (don't exactly remember) to accomplish this.
However, is there any chance value_group is null allowed? Tom Kyte in his book shows this influences the way NOT IN subqueries are handled, this might also apply to IN subqueries. Guess you'd better post the full plan, and information on the indexes present on the main table.

Usually, btw, transforming an IN subquery to an EXISTS correlated subquery still works.

--
Sybrand Bakker, Senior Oracle DBA
Received on Wed Feb 09 2005 - 12:27:50 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US