[postgis-tickets] [PostGIS] #4750: Performance issue with computed columns

[PostGIS]

#4750: Performance issue with computed columns
----------------------+---------------------------
  Reporter:  robe     |      Owner:  pramsey
      Type:  defect   |     Status:  new
  Priority:  medium   |  Milestone:  PostGIS 3.0.3
 Component:  postgis  |    Version:  3.0.x
Resolution:           |   Keywords:
----------------------+---------------------------

Comment (by Algunenano):

 Note that in the fast plan each parallel worker (2 of them) is doing a seq
 scan over co_random_points and then for each point (10k each) an index
 scan over over ix_co_road_line_way_gist. In the slow plan it's happening
 the other way around, each worker does a seq scan over co_road_line, and
 then for each line (666688 each) it does a seq scan over the points.

 So it looks that the thing that's making things slow is not having good
 stats and once that is solves things work as expected. Can you try to do
 the process analyzing both tables (the random points and the ones with the
 generated column) before running any query and see if those times are
 normal?

-- 
Ticket URL: <https://trac.osgeo.org/postgis/ticket/4750#comment:6>
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.