[postgis-devel] RECHECK condition on geometry index scans

Paragon Corporation lr at pcorp.us
Mon Feb 24 03:49:59 PST 2014


http://www.postgresql.org/message-id/464F3C5D.2000700@enterprisedb.com

"The Recheck condition is always shown, even if the bitmap is not stored 
as lossy and no rechecking is done."
 

-----Original Message-----
From: postgis-devel-bounces at lists.osgeo.org
[mailto:postgis-devel-bounces at lists.osgeo.org] On Behalf Of Sandro Santilli
Sent: Monday, February 24, 2014 2:29 AM
To: Paul Ramsey
Cc: postgis-devel at lists.osgeo.org
Subject: Re: [postgis-devel] RECHECK condition on geometry index scans

On Fri, Feb 21, 2014 at 07:03:20AM -0800, Paul Ramsey wrote:
> Dunno how you're still seeing it, could it be an artifact of PgSQL
explain? Like, the core system says "I wonder if I need to recheck" and then
after asking us (and we say "no") it moves on, but the initial question
still shows up in the explain output? We definitely say "no" in both the 2d
and nd index bindings.

Spotted on a PostgreSQL 9.3.0 with a PostGIS lib "2.1.1 r12113"
and procs "2.1.0 r11822". 

The recheck seems to be only in the bitmap heap scan indeed:

  ->  Bitmap Heap Scan on tab  (cost=372.82..28794.17 rows=5441 width=4)
        Recheck Cond: ((rast)::geometry && '...'::geometry)
        Filter: _st_intersects('...'::geometry, rast, NULL::integer)
        ->  Bitmap Index Scan on tab_rast_gist_idx  (cost=0.00..372.55
rows=16322 width=0)
              Index Cond: ((rast)::geometry && '...'::geometry)

Do we have any say on that one ?

--strk;
_______________________________________________
postgis-devel mailing list
postgis-devel at lists.osgeo.org
http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-devel





More information about the postgis-devel mailing list