Re: BIGINT indexes still with problems - Mailing list pgsql-general

From Peter Eisentraut
Subject Re: BIGINT indexes still with problems
Date
Msg-id 200408102125.53737.peter_e@gmx.net
Whole thread Raw
In response to BIGINT indexes still with problems  ("Dan Ruthers" <dan211a@lycos.com>)
List pgsql-general
Dan Ruthers wrote:
> The index is used!
> I also did a vacuum analyze, and restarted Postgres and it did not
> make any difference. I tried many other ID values (ex 783218 and
> 783220), and they seem to use the index correctly. Only that value
> doesn't.

Possibly, that is the most common value and the cost calculation yields
that it would be more efficient to not use the index.  If you disagree,
please show the timings generated by EXPLAIN ANALYZE.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/


pgsql-general by date:

Previous
From: Sven Willenberger
Date:
Subject: pg_restore with Fc fails with [archiver] out of memory error
Next
From: Gaetano Mendola
Date:
Subject: Re: PostgreSQL 8.0.0 Officially Goes Beta