Русский

Re: Very specialised query

thread

List

Post date

Sort by

From:	Matthew Wakeling
Subject:	Re: Very specialised query
Date:	March 30, 2009 , 10:28:30
Msg-id:	alpine.DEB.2.00.0903301101410.21772@aragorn.flymine.org (view: Whole thread, Raw)
In response to:	Re: Very specialised query ("Marc Mamin")
Responses:	Re: Very specialised query ("Marc Mamin")
List:	pgsql-performance

Tree view

Tree view

Very specialised query (Matthew Wakeling, 26 March 2009, 14:31:09 )
Re: Very specialised query ("Kevin Grittner", 26 March 2009, 14:47:44 )
Re: Very specialised query (Tom Lane, 26 March 2009, 15:50:35 )
Re: Very specialised query (Matthew Wakeling, 26 March 2009, 16:06:07 )
Re: Very specialised query (Matthew Wakeling, 27 March 2009, 12:46:00 )
Re: Very specialised query (Tom Lane, 27 March 2009, 14:30:13 )
Re: Very specialised query (Віталій Тимчишин, 27 March 2009, 16:36:55 )
Re: Very specialised query (Matthew Wakeling, 27 March 2009, 17:34:36 )
Re: Very specialised query (Tom Lane, 27 March 2009, 18:43:27 )
Re: Very specialised query (Matthew Wakeling, 30 March 2009, 10:24:41 )
Re: Very specialised query (Matthew Wakeling, 30 March 2009, 15:57:23 )
Re: Very specialised query (Віталій Тимчишин, 30 March 2009, 17:29:20 )
Re: Very specialised query (Matthew Wakeling, 31 March 2009, 17:08:04 )
Re: Very specialised query (Віталій Тимчишин, 31 March 2009, 21:12:02 )
Re: Very specialised query (Matthew Wakeling, 01 April 2009, 17:10:49 )
Re: Very specialised query (Dimitri Fontaine, 27 March 2009, 20:08:43 )
Re: Very specialised query (Matthew Wakeling, 30 March 2009, 10:26:16 )
Re: Very specialised query ("Marc Mamin", 27 March 2009, 22:55:30 )
Re: Very specialised query (Matthew Wakeling, 30 March 2009, 10:28:30 )
Re: Very specialised query ("Marc Mamin", 30 March 2009, 15:36:12 )
Re: Very specialised query (Matthew Wakeling, 30 March 2009, 15:59:11 )
Re: Very specialised query (Matthew Wakeling, 30 March 2009, 16:22:18 )
Re: Very specialised query (Віталій Тимчишин, 30 March 2009, 18:15:45 )
Re: Very specialised query (Matthew Wakeling, 01 April 2009, 17:33:21 )
Re: Very specialised query (Matthew Wakeling, 01 April 2009, 18:27:25 )
Re: Very specialised query (Matthew Wakeling, 02 April 2009, 15:02:35 )
Re: Very specialised query (Craig Ringer, 02 April 2009, 16:36:15 )
Re: Very specialised query ("Marc Mamin", 30 March 2009, 17:48:38 )
Re: Very specialised query (Matthew Wakeling, 31 March 2009, 11:54:52 )

On Fri, 27 Mar 2009, Marc Mamin wrote:
> if your data are mostly static and you have a few mains objects,
> maybe you can have some gain while defining conditional indexes for those plus one for the rest
> and then slicing the query:

Maybe. I thought about doing that. However, I am not convinced that would
be much of a gain, and would require major rewriting of the queries, as
you suggest.

> WHERE (l2.start BETWEEN  l1.start AND l1.end
>          OR
>          l1.start BETWEEN  l2.start AND l2.end
>          )

Yes, that's another way to calculate an overlap. However, it turns out to
not be that fast. The problem is that OR there, which causes a bitmap
index scan, as the leaf of a nested loop join, which can be rather slow.

Matthew

--
 I'd try being be a pessimist, but it probably wouldn't work anyway.

pgsql-performance by date:

From: Matthew Wakeling
Date: 30 March 2009, 10:22:18
Subject: Re: Very specialised query

From: Scott Marlowe
Date: 30 March 2009, 10:33:39
Subject: Re: Forcing seq_scan off for large table joined with tiny table yeilds improved performance