Re: performace problem after VACUUM ANALYZE

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Scott Cain <cain(at)cshl(dot)org>
Cc: pgsql-performance(at)postgresql(dot)org, gmod schema <gmod-schema(at)lists(dot)sourceforge(dot)net>
Subject: Re: performace problem after VACUUM ANALYZE
Date: 2003-02-14 17:00:18
Message-ID: 8000.1045242018@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Scott Cain <cain(at)cshl(dot)org> writes:
> [ much stuff ]

Could we see EXPLAIN ANALYZE, not just EXPLAIN, output for all these
alternatives? Your question boils down to "why is the planner
misestimating these queries" ... which is a difficult question to
answer when given only the estimates and not the reality.

A suggestion though is that you might need to raise the statistics
target on the indexed columns, so that ANALYZE will collect
finer-grained statistics. (See ALTER TABLE ... SET STATISTICS.)
Try booting it up to 100 (from the default 10), re-analyze, and
then see if/how the plans change.

regards, tom lane

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Scott Cain 2003-02-14 17:29:36 Re: performace problem after VACUUM ANALYZE
Previous Message Scott Cain 2003-02-14 16:44:00 performace problem after VACUUM ANALYZE