pg.ddx.io  pgsql-performance@postgresql.org mailing list archive  
help / color / mirror / Atom feed
From: Jim C. Nasby <jnasby@pervasive.com>
To: PFC <lists@peufeu.com>
Cc: Martijn van Oosterhout <kleptog@svana.org>
Cc: Christian Kratzer <ck@cksoft.de>
Cc: pgsql-performance@postgresql.org
Cc: pgsql-hackers@postgresql.org
Subject: Re: [HACKERS] Big IN() clauses etc : feature proposal
Date: Wed, 10 May 2006 14:06:17 -0500
Message-ID: <20060510190617.GP99570@pervasive.com> (raw)
In-Reply-To: <op.s897j60qcigqcu@apollo13>
References: <1147110142.32369.7.camel@archimedes>
	<20060508175013.GY99570@pervasive.com>
	<28288.1147111196@sss.pgh.pa.us>
	<op.s89zl30wcigqcu@apollo13>
	<20060509105214.P90693@vesihiisi.cksoft.de>
	<op.s8916gm6cigqcu@apollo13>
	<20060509113337.M90693@vesihiisi.cksoft.de>
	<op.s893vzcpcigqcu@apollo13>
	<20060509103632.GD29652@svana.org>
	<op.s897j60qcigqcu@apollo13>

On Tue, May 09, 2006 at 01:29:56PM +0200, PFC wrote:
> 0.101 ms BEGIN
> 1.451 ms CREATE TEMPORARY TABLE tmp ( a INTEGER NOT NULL, b INTEGER NOT  
> NULL, c TIMESTAMP NOT NULL, d INTEGER NOT NULL ) ON COMMIT DROP
> 0.450 ms INSERT INTO tmp SELECT * FROM bookmarks ORDER BY annonce_id DESC  
> LIMIT 20
> 0.443 ms ANALYZE tmp
> 0.365 ms SELECT * FROM tmp
> 0.310 ms DROP TABLE tmp
> 32.918 ms COMMIT
> 
> 	CREATING the table is OK, but what happens on COMMIT ? I hear the 
> 	disk  seeking frantically.
> 
> With fsync=off, I get this :
> 
> 0.090 ms BEGIN
> 1.103 ms CREATE TEMPORARY TABLE tmp ( a INTEGER NOT NULL, b INTEGER NOT  
> NULL, c TIMESTAMP NOT NULL, d INTEGER NOT NULL ) ON COMMIT DROP
> 0.439 ms INSERT INTO tmp SELECT * FROM bookmarks ORDER BY annonce_id DESC  
> LIMIT 20
> 0.528 ms ANALYZE tmp
> 0.364 ms SELECT * FROM tmp
> 0.313 ms DROP TABLE tmp
> 0.688 ms COMMIT
> 
> 	Getting closer ?
> 	I'm betting on system catalogs updates. I get the same timings with  
> ROLLBACK instead of COMMIT. Temp tables have a row in pg_class...

Have you tried getting a profile of what exactly PostgreSQL is doing
that takes so long when creating a temp table?

BTW, I suspect catalogs might be the answer, which is why Oracle has you
define a temp table once (which does all the work of putting it in the
catalog) and then you just use it accordingly in each individual
session.
-- 
Jim C. Nasby, Sr. Engineering Consultant      jnasby@pervasive.com
Pervasive Software      http://pervasive.com    work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf       cell: 512-569-9461



view thread (44+ messages)  latest in thread

Message-ID: <20060510190617.GP99570@pervasive.com>
Permalink:  ../20060510190617.GP99570@pervasive.com/
Also on:    postgresql.org/message-id/20060510190617.GP99570@pervasive.com

 · 

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: pgsql-performance@postgresql.org
  Cc: jnasby@pervasive.com, lists@peufeu.com, kleptog@svana.org, ck@cksoft.de, pgsql-hackers@postgresql.org
  Subject: Re: [HACKERS] Big IN() clauses etc : feature proposal
  In-Reply-To: <20060510190617.GP99570@pervasive.com>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox