pg.ddx.io pgsql-sql@postgresql.org mailing list archivehelp / color / mirror / Atom feed
(no subject) 2+ messages / 2 participants [nested] [flat]
* (no subject) @ 2006-08-29 16:44 Sumeet <asumeet@gmail.com> 0 siblings, 1 reply; 2+ messages in thread From: Sumeet @ 2006-08-29 16:44 UTC (permalink / raw) To: pgsql-sql Hi All, I'm having very strange problems with indexing, I have a large number of partial indexes, when i try to run some queries sometimes the queries run very fast and sometimes they are 100times slower than what i ran before.....I tried running vacuum analyze couple of times and it dint help me at all. Also i tried the explain analyze and what i found was the query execution plan keeps on changes. Is there any better way to gain control over these, I tried turning seqscan off, and messing with some other parameters but dint really help. Also can somebody point me out to links which explains more about query execution plans..i.e. each of parameters like bitmap heap scan, index scan....etc... -- Thanks, Sumeet ^ permalink raw reply [nested|flat] 2+ messages in thread
* Re: @ 2006-08-29 17:40 Andrew Sullivan <ajs@crankycanuck.ca> parent: Sumeet <asumeet@gmail.com> 0 siblings, 0 replies; 2+ messages in thread From: Andrew Sullivan @ 2006-08-29 17:40 UTC (permalink / raw) To: pgsql-sql On Tue, Aug 29, 2006 at 12:44:28PM -0400, Sumeet wrote: > me at all. Also i tried the explain analyze and what i found was the query > execution plan keeps on changes. Is there any better way to gain control My bet is that you don't have enough sample data to support a consistent query plan. SET STATISTICS might be what you need. > parameters but dint really help. Also can somebody point me out to links > which explains more about query execution plans..i.e. each of parameters > like bitmap heap scan, index scan....etc... On [mumble techdocs I think] there's a tutorial called EXPLAIN explained. Google will help you -- just use that title, with maybe "postgresql" tacked onto your search string. A -- Andrew Sullivan | ajs@crankycanuck.ca The plural of anecdote is not data. --Roger Brinner ^ permalink raw reply [nested|flat] 2+ messages in thread
end of thread, other threads:[~2006-08-29 17:40 UTC | newest] Thread overview: 2+ messages (download: mbox mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2006-08-29 16:44 Sumeet <asumeet@gmail.com> 2006-08-29 17:40 ` Re: Andrew Sullivan <ajs@crankycanuck.ca>
This inbox is served by DDX for PostgreSQL; see mirroring instructions for how to clone and mirror all data and code used for this inbox