Re: CBO influences

"DA Morgan" <damorgan_at_psoug.org> wrote in message news:1126329859.958172_at_yasure...
>
> One more, slightly related question, if I may.
>
> Gathering system statistics I do not seem to be able to force
> a demo database to gather statistics for MBRC, MREADTIM, and
> SLAVETHR.
>
> I understand the last might be a bit difficult on a notebook
> used for classroom demo. But any help on any of this would
> be appreciated.
>
> Thanks.
> --
> Daniel A. Morgan
> http://www.psoug.org
> damorgan_at_x.washington.edu
> (replace x with u to respond)

Perhaps your sample data sets and
are too small, and your db_cache_size
too big to generate any physical
multiblock reads. I just create a table that I know is larger than the cache,
then do a few count(*) on it.

(Or if the cache is small, I do a few
scans of source$ and a few forced
indexed accesses of source$)

The SLAVETHR is a bit of a puzzle,
since you haven't mentioned MAXTHR
as not being set - and these both apply
to parallel queries: MAXTHR is the
observation of throughput the coordinator can take, SLAVETHR is the throughput
the slaves can supply - I wouldn't have
thought you could get one set without
the other. (Guess - maybe the gather
goes wrong if you have had only one
ongoing parallel query that didn't terminate when you are trying to gather system stats).

-- 
Regards

Jonathan Lewis

http://www.jlcomp.demon.co.uk/cbo_book/ind_book.html
Cost Based Oracle - Volume 1: Fundamentals
On-shelf date: Nov 2005

http://www.jlcomp.demon.co.uk/faq/ind_faq.html
The Co-operative Oracle Users' FAQ

http://www.jlcomp.demon.co.uk/appearances.html
Public Appearances - schedule updated 4th Sept 2005