[postgis-users] Indexes (ENABLE_SEQSCAN)

[Paul Ramsey]

Run "select update_geometry_stats();" and see if the planner does not 
get smarter about whether or not to use the index.

P.

Alexandre Florio wrote:
> Analyzing the archives, this seems to be an exhaustively discussed
> problem which was apparently fixed in version 0.6.
> 
> The documentation says that I do not need, anymore, to use "SET
> ENABLE_SEQSCAN=OFF" to force the planner to use the spatial index.
> Nevertheless, I have a database in which a query do execute much
> faster when setting ENABLE_SEQSCAN to OFF.
> 
> Before dumping the table scheme, queries and outputs I would like to
> confirm that this should not happen in the current version of PostGIS
> (0.8.2).
> 
> Thanks, Alexandre Florio