Re: Poorly performing query and 10053 trace
- From: Jurijs Velikanovs <j.velikanovs@xxxxxxxxx>
- To: Boris Dali <boris_dali@xxxxxxxx>
- Date: Thu, 12 Jan 2006 17:54:51 -0800
Ups :)
On 1/12/06, Boris Dali <boris_dali@xxxxxxxx> wrote:
>
> Wasn't it the opposite - if at least one table
> analyzed in a join, than the cbo kicks in and made-up
> stats used for non-analyzed tables?
>
> Thanks,
> Boris Dali.
>
> --- Jurijs Velikanovs <j.velikanovs@xxxxxxxxx> wrote:
>
> > >> this query accesses 2 analyzed tables
> > >> and 3 unanalyzed tables
> > If I recall correctly in 817 if some of the tables
> > have no statistics
> > Oracle uses RULE based optimizer.
> > For RULE based optimizer the order in which indexes
> > have been created
> > some times play trumendios role.
> > Can you confirm the that query parsed in the RULE
> > mode?
> >
> > Jurijs
>
>
>
>
>
>
>
>
> __________________________________________________________
> Find your next car at http://autos.yahoo.ca
>
--
Jurijs
+44 7738 013090 (GMT)
============================================
http://otn.oracle.com/ocm/jvelikanovs.html
--
//www.freelists.org/webpage/oracle-l
Other related posts:
