Since the statistics are gathered by running undocumented SQL, and any SQL
is itself subject to the mercy of initialization parameters, for example
db_multiblock_read_count for the full table scans, then I would say yes...
just none that I can think of in 8... The modelling in 9 and 10 ... TYPICAL
that kind of thing would affect it though.
Telemachus.... nice to be back !
"Niall Litchfield" <niall.litchfield_at_dial.pipex.com> wrote in message
news:41976743$0$16437$cc9e4d1f_at_news-text.dial.pipex.com...
> "DA Morgan" <damorgan_at_x.washington.edu> wrote in message
> news:1100394206.441584_at_yasure...
> > wagen wrote:
> >
> >> We gather statistics for the database objects (user created) using
> >> DBMS_STATS (9i +), ANALYZE (8i +).
> >>
> >> Do ANY of the initialization parameters influence the way the
> >> statistics are gathered? If YES, what are the init.ora parameters?
> >>
> >> NOTE: I am not referring to ESTIMATE/COMPUTE/SAMPLE...
> >>
> >> Thanks for any pointers.
> >>
> >> Wagen
> >
> > I believe the answer is ... gathered no ... utilized by the
> > optimizer yes.
>
> I expect that setting the PARALLEL_XXX parameters to enable query
> parallelism at different degrees and then setting the degree parameters to
> DBMS_STATS differently would result in a difference in statitics
gathering.
> I haven't tested this though.
>
> DBMS_STATS works pretty well in 8i as well BTW.
>
>
> --
> Niall Litchfield
> Oracle DBA
> http://www.niall.litchfield.dial.pipex.com
>
>
Received on Mon Nov 15 2004 - 11:56:36 CST
