[postgis-users] ST_Segmentize(geography) perfomance?

Paragon Corporation lr at pcorp.us
Sat Dec 8 06:08:43 PST 2012

That's the best time to start finding performance issues, before we release.
So keep it coming.
As far as reporting these, please post a ticket to trac
with Milestonre 2.1.0 and the
SELECT version();
SELECT postgis_full_version();
of your install.
More importantly a sample data output e.g. geometry wkt or a table with
single record that is really slow.
If it's a performance issue, make sure to note the timing you are getting.
realityexists is really good at posting bug reports, so his is probably a
good model to follow:  Here is an example of his


From: postgis-users-bounces at lists.osgeo.org
[mailto:postgis-users-bounces at lists.osgeo.org] On Behalf Of Shaun Walbridge
Sent: Friday, December 07, 2012 10:04 PM
To: postgis-users at lists.osgeo.org
Subject: [postgis-users] ST_Segmentize(geography) perfomance?


I'd like to use the bleeding-edge feature of ST_Segmentize on a collection
of about 20,000 multilinestring geometries, each containing 1000-50000
vertices. For most of my geometries, things went without a hitch, but a
couple of subtables seem to just 'hang' when trying to segmentize. I've
mucked around with my shared_buffers to make sure that the geometries can
fit into memory, but it doesn't seem to have helped. My query looks like

CREATE TABLE arcs_input AS SELECT id, name, count, ST_Segmentize(geog,
10000) as loc FROM lines_input;

I know this is still only in SVN, but what's the best way for me to help
debug this issue?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-users/attachments/20121208/e092a1a1/attachment.html>

More information about the postgis-users mailing list