Re: Determining scan types - Mailing list pgsql-general

From Philip Molter
Subject Re: Determining scan types
Date
Msg-id 20010703121831.V12723@datafoundry.net
Whole thread Raw
In response to Re: Determining scan types  ("Richard Huxton" <dev@archonet.com>)
List pgsql-general
On Tue, Jul 03, 2001 at 05:12:43PM +0100, Richard Huxton wrote:
: VACUUM ANALYZE frequency depends on numbers of updates. I believe someone
: has been looking at a way of doing this in the background.

Oh yeah, definitely depends on updates, or rather, changes to the
table contents (insertions, deletions).

: For the purposes of setting SEQSCAN try something like:
:
: SET ENABLE_SEQSCAN TO OFF;
:
: Can't remember if it applies to this transaction or this connection. Run a
: grep on the docs - you'll only find a couple of hits.

Well, I turned it off for the entire database (since right now, we're
only using the db for this one application), but I lose the benefit of
seqscans in situations where it's appropriate.  That's why I was
wondering if there's anyway to tell the optimizer to prefer index scans
over sequential scans when it has a choice.  Right now, it's using less
efficient joining methods where it would normally use sequential scans
(not that I'm complaining too much; CPU performance has more than
doubled since taking out sequential scans).

* Philip Molter
* DataFoundry.net
* http://www.datafoundry.net/
* philip@datafoundry.net

pgsql-general by date:

Previous
From: "Richard Huxton"
Date:
Subject: Re: Determining scan types
Next
From: Tom Lane
Date:
Subject: Re: pqReadData() -- backend closed on COPY