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

Home -> Community -> Mailing Lists -> Oracle-L -> Re: Optimization of Partitioned Outer Joins

Re: Optimization of Partitioned Outer Joins

From: jaromir nemec <jaromir_at_db-nemec.com>
Date: Mon, 3 Jan 2005 22:37:51 +0100
Message-ID: <0e8101c4f1dc$7b7b6500$3c02a8c0@JARAWIN>


Hi,

> but the original times table is
> the preserved table in the outer join, so for a
> nested loop it would be the outer table.

The preserved table is equal the outer table for SM outer join and for hash outer join as well, isn't it?
I'm little bit confused from the formulation above, based on the nested loop context.

By the way, what is the correct definition of an outer table in an (inner) hash join? (There is no preserved table to be used in the definition; in case that ORDERED hint isn't used, it is not obvious what table is hashed and what table is probed).

The definition of "outer table" is particularly important e.g. for the hint PQ_DISTRIBUTE.
Can anybody clarify it for me?

thanks

Jaromir

--
http://www.freelists.org/webpage/oracle-l
Received on Mon Jan 03 2005 - 15:40:34 CST

Original text of this message

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