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: tough choices

Re: tough choices

From: Mark Townsend <markbtownsend_at_comcast.net>
Date: Thu, 24 Jun 2004 03:22:37 GMT
Message-ID: <0OrCc.166484$Ly.102964@attbi_s01>


Serge Rielau wrote:

> In a "partitioned" view case, which is the one we are debating here,
> The compiler will find that the same indices are present for all tables.
> The SELECT FROM V WHERE <input> with V = UNION ALL (T1, ..., Tn) will be
> rewritten into a SELECT FROM T(i) with i = foo(<input>) and associated
> indexes Ix(i)

For secondary indexes as well ? Can uniqueness be enforced across the tables for things other than the primary key ? What happens if a constraint is not in place at the time of the DML ?

> Statistics are also computed with some magic.

Including overall skew at the table or index level ? Will one or more individual execution plans be merged into an overall range scan and executed in parallel if it makes sense ? Does the possible degree of parallelism only match the tables left in the query after the elimination. Received on Wed Jun 23 2004 - 22:22:37 CDT

Original text of this message

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