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: Explain Plan For Query

Re: Explain Plan For Query

From: Niall Litchfield <n-litchfield_at_audit-commission.gov.uk>
Date: Sat, 13 Apr 2002 10:46:26 +0100
Message-ID: <3cb7fe51$0$234$ed9e5944@reading.news.pipex.net>


Clarification below
"Howard J. Rogers" <dba_at_hjrdba.com> wrote in message news:a9706b$3cb$1_at_lust.ihug.co.nz...
> It doesn't necessarily "favour" full scans. It just means the plan is
> optimised for returning the complete resultset, rather than for getting
the
> quickest possible return of the first part of a resultset (first_rows).
All
> that means is that the optimizer will spend longer choosing the perfect
> plan.

<snip>> "Niall Litchfield" <n-litchfield_at_audit-commission.gov.uk> wrote in message

> > > ALL_ROWS
> > > retrieves the complete resultset in SGA
> > > favors full tablescans.
> >
> > But I didn't know that. Is this difference documented/demonstrated
> > somewhere?

Actually I was trying to comment on the statement that the complete resultset was assembled in memory (rather than the FTS comment), presumably before passing back to the user.

Is it the case that the optimiser considers more plans for all_rows or just that the goal is different minimum execution time not maximum response time?

--
Niall Litchfield
Oracle DBA
Audit Commission UK
*****************************************
Please include version and platform
and SQL where applicable
It makes life easier and increases the
likelihood of a good answer

******************************************
Received on Sat Apr 13 2002 - 04:46:26 CDT

Original text of this message

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