| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | "Jim C(dot) Nasby" <jnasby(at)pervasive(dot)com> |
| Cc: | Martijn van Oosterhout <kleptog(at)svana(dot)org>, "Florian G(dot) Pflug" <fgp(at)phlo(dot)org>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: An Idea for planner hints |
| Date: | 2006-08-09 19:26:59 |
| Message-ID: | 28513.1155151619@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Jim C. Nasby" <jnasby(at)pervasive(dot)com> writes:
> On Wed, Aug 09, 2006 at 02:02:10PM +0200, Martijn van Oosterhout wrote:
>> Perhaps the way to go would be to allow users to declare columns often
>> used together and have ANALYSE collect information on correlation which
>> can be used later...
> One thing that would help tremendously would be to collect stats on
> multi-column indexes. That would probably hit a good chunk of our
> problem areas.
But it would specifically fail to cover join situations. I kinda like
Martijn's thought of allowing users to specify combinations of columns
to collect correlation stats about.
(Not sure how we'd implement that, seeing that ANALYZE currently works
on one table at a time, but it's probably doable --- and it'd fix the
fundamental problem for correlation statistics, which is how not to try
to collect stats about an exponential number of combinations ...)
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Shoaib Mir | 2006-08-09 19:30:10 | Re: WIN32 Build? |
| Previous Message | Mark Dilger | 2006-08-09 19:26:15 | Re: An Idea for planner hints |