Skip site navigation (1) Skip section navigation (2)

RE: Use of index in 7.0 vs 6.5

From: "Hiroshi Inoue" <Inoue(at)tpf(dot)co(dot)jp>
To: "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "Ryan Bradetich" <ryan_bradetich(at)hp(dot)com>, <pgsql-sql(at)postgresql(dot)org>
Subject: RE: Use of index in 7.0 vs 6.5
Date: 2000-05-25 07:19:54
Message-ID: 000e01bfc619$9fa5f9c0$ (view raw, whole thread or download thread mbox)
Lists: pgsql-sql
> -----Original Message-----
> From: pgsql-sql-owner(at)hub(dot)org [mailto:pgsql-sql-owner(at)hub(dot)org]On Behalf
> Of Tom Lane

> This seems to point up (once again) the deficiency of assuming that
> the most-common value in the table is a good guide to the frequency
> of typical values.  You showed that host_id = 446 occurs in 2.8% of
> the rows in this table; a search for 446 very probably would be faster
> as a seqscan than as an indexscan (you might care to try it and see).
> But that's probably a statistical outlier that's not got much to do
> with the frequency of typical values in the table.
> The only really good answer to this problem is to collect more-detailed
> statistics in VACUUM ANALYZE, which I hope to see us doing in a release
> or so.

For example we could count up distinct values for the first column of an
index by scanning its index relation.
> In the meantime I am wondering about deliberately skewing the
> cost model in favor of indexscans, because I sure haven't heard many
> complaints about erroneous selection of indexscans...
> One way to put a thumb on the scales is to reduce the value of the SET
> variable random_page_cost.  The default value is 4.0, which seems to
> correspond more or less to reality, but reducing it to 3 or so would
> shift the planner pretty nicely in the direction of indexscans.

Or how about changing current fudge factor ?
For example,from 0.5 to 0.2 which is the fudge factor of attdisbursion


Hiroshi Inoue

In response to


pgsql-sql by date

Next:From: Mikheev, VadimDate: 2000-05-25 14:29:58
Subject: RE: Use of index in 7.0 vs 6.5
Previous:From: Tom LaneDate: 2000-05-25 06:14:38
Subject: Re: Use of index in 7.0 vs 6.5

Privacy Policy | About PostgreSQL
Copyright © 1996-2017 The PostgreSQL Global Development Group