[postgis-users] Partitioning spatial table

Bernhard Reimar Hoefle Bernhard.Hoefle at uibk.ac.at
Wed Feb 22 05:54:37 PST 2006


> Message: 4
> Date: Tue, 21 Feb 2006 12:32:56 +0100
> From: Arnaud Lesauvage <thewild at freesurf.fr>
> Subject: Re: [postgis-users] Partitioning spatial table
> To: PostGIS Users Discussion <postgis-users at postgis.refractions.net>
> Message-ID: <43FAFA68.70102 at freesurf.fr>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Hi all !
>
> Sorry for not giving more feedback, I have been quite busy.
>
> I tried Bill's suggestion and it works great !
> Basically, here is what I did :
> I created a generic table.
> I created two child tables with the exact same structure (I have 2
> datasets from two countries, so I thought it might be a good idea
> to split them).
> I created the index on the integer field for both child tables,
> and clustered them on this index.
> I then created the conditional GiST geometry indexes on the child
> tables, with the integer field as a condition.
>
> It works very well ! The only thing that slows down the query is
> the scan to read the data, but since almost all the dataset has to
> be retrieved, I believe there is no way to speed this up (appart
> from having faster disks).
>
> I haven't tried the multicolumn indexes yet.
> Am I right to assume that since the tables are clustered on the
> integer field, it would be useless to configure a multicolumn
> index (integer, geometry) ?
>
> Regards
> --
> Arnaud

Hi Arnaud!
Could you please post your SQL statements for the tables and conditional GiST
indexes.

Bernhard



More information about the postgis-users mailing list