Re: Performance on temp table inserts - Mailing list pgsql-sql

From Josh Berkus
Subject Re: Performance on temp table inserts
Date
Msg-id 200305191504.58368.josh@agliodbs.com
Whole thread Raw
In response to Performance on temp table inserts  (Jeff Boes <jboes@nexcerpt.com>)
List pgsql-sql
Jeff,

> Box #1 is our "production" server, and actually has more going on during
> the day than #2 which is the "test" server.  #1 has "shared_buffers" set
> to 131072; #2 has it at 250000. Hardware, O/S and PostgreSQL versions are
> all identical.  Software executing the query (Perl code using DBI) is the
> same.

First, I think you need to join the PGSQL-PERFORMANCE list; we discuss things
like your issue all the time there.

Collective wisdom on that list is that shared_buffer settings above 5000-6000
actually degrade performance by robbing the kernel buffer of resources.  The
postgres shared_buffers are just a "holding area" for operations spooling to
the kernel buffer which actually does most of the work.

HTH.

--
-Josh BerkusAglio Database SolutionsSan Francisco



pgsql-sql by date:

Previous
From: "James Taylor"
Date:
Subject: CASE
Next
From: Stephan Szabo
Date:
Subject: Re: "deadlock detected" / cascading locks