[postgis-tickets] [PostGIS] #3753: Gist penalty misbehaves if points are inserted in gridded ordered fashion

PostGIS trac at osgeo.org
Sun Sep 3 02:30:26 PDT 2017


#3753: Gist penalty misbehaves if points are inserted in gridded ordered fashion
----------------------+---------------------------
  Reporter:  komzpa   |      Owner:  pramsey
      Type:  defect   |     Status:  new
  Priority:  blocker  |  Milestone:  PostGIS 2.4.0
 Component:  postgis  |    Version:  2.3.x
Resolution:           |   Keywords:
----------------------+---------------------------

Comment (by x4m):

 Sorry, I was a bit busy these days.

 Replying to [comment:14 robe]:
 > So I think all that is left before we can close this ticket is to
 perhaps drop the lines you referred to and some retesting to confirm it
 improves things.

 Previously I observed following
 >Before patch selection takes 300 seconds (on my machine) after patch
 selection takes 37 seconds... 2D data in 3D

 >integer grid for 2D ... index with 43MB size and 109.2 seconds for
 select. With patch index size is 56MB and select time 68.2 seconds

 This constitute statistically significant performance improvement. Do you
 want to restart these tests again?

 As to
 >drop the lines you referred to
 In May, we have made tests with @Komzpa and found no improvement in
 dropping the lines. Sorry for not putting that information to the
 discussion, we were pinging pictures and gevel images to spot the
 difference(induced by dropping the lines) , but found none.

 From my point of view, current state is:
 1. Best we can achieve by penalty function tuning
 2. Contains sufficient improvement over revision 15391

--
Ticket URL: <https://trac.osgeo.org/postgis/ticket/3753#comment:16>
PostGIS <http://trac.osgeo.org/postgis/>
The PostGIS Trac is used for bug, enhancement & task tracking, a user and developer wiki, and a view into the subversion code repository of PostGIS project.


More information about the postgis-tickets mailing list