thread-index: Acc2gyzWjf51C8K/QL+EGDfdhHCmMg==
Content-Class: urn:content-classes:message
Importance: normal
Subject: Re: [HACKERS] table partioning performance
From: "Simon Riggs" <simon@2ndquadrant.com>
To: "Jim C. Nasby" <jim@nasby.net>
Cc: "Steven Flatt" <steven.flatt@gmail.com>,
	"Colin Taylor" <colin.taylor@gmail.com>, <pgsql-performance@postgresql.org>
In-Reply-To: <20070111210122.GR36267@nasby.net>
References: <84d3067d0701062037k3229c9ddjbde73b1fd79ee209@mail.gmail.com>
	<357fa7590701081202p726c049ej61390dd1816d1369@mail.gmail.com>
	<1168346933.3951.276.camel@silverbirch.site>
	<357fa7590701101300m2812545k1487ca58e97970ed@mail.gmail.com>
	<1168517751.3951.538.camel@silverbirch.site>
	<20070111210122.GR36267@nasby.net>
Content-Type: text/plain;
	charset="iso-8859-1"
Date: Fri, 12 Jan 2007 19:50:54 +0000
Message-ID: <1168631454.3990.129.camel@silverbirch.site>
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit

On Thu, 2007-01-11 at 15:01 -0600, Jim C. Nasby wrote:

> I'm wondering if you see any issues with COPYing into a partitioned
> table that's using triggers instead of rules to direct data to the
> appropriate tables?

The data demographics usually guides you towards what to do.

You could COPY into the table that would receive most rows and use
before triggers to INSERT into the other tables, rather than the main
one. I'd be surprised if that was very fast for an even distribution
though. It could well be faster if you have a small number of rows into
a large number of targets because that would be quicker than re-scanning
a temp table repeatedly just to extract a few rows each time.

-- 
  Simon Riggs             
  EnterpriseDB   http://www.enterprisedb.com