[postgis-devel] Benchmark/Testing setup

Brian M Hamlin maplabs at light42.com
Tue Feb 19 10:14:15 PST 2019


 
Hi Bruce -
 
  Thanks for this opportunity to look more carefully at a crucial topic .. 
Please document the BASE_OS, PostgreSQL_VERSION, postgres conf 
settings, versions of dependent libs and
hardware involved .. tablespace physical disk, RAM, CPU model..  Its a 
bit of work, but important !
 
  from that base, the PostGIS BRANCH and VERSION can be varied. 
Perhaps PostgreSQL itself as well. 
The values there might be recorded in some tabular format, including 
the results -> per test suite. 
 
You probably know about the fine site   
https://explain.depesz.com/    which can add detail to your EXPLAIN 
findings. 
 
  best regards from Berkeley, California
   --Brian
 

On Tue, 19 Feb 2019 09:59:12 -0800, Bruce Rindahl
  wrote:

       I now have my cluster setup for testing with multiple small 
computers with a full PostgreSQL/PostGIS installation.  Using this I 
ran some quick tests to find what cost on the ST_Area function kicks 
into parallel mode.  Using data/queries by Paul Ramsey 
at:http://blog.cleverelephant.ca/2017/11/parallel-postgis-2A.html
It kicks into parallel mode at cost = 66.  I can easily test other 
functions/configuration. 
 If this can help, let me knowThanks
Bruce

-------------------------

_______________________________________________
postgis-devel mailing list
postgis-devel at lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/postgis-devel

--
Brian M Hamlin
OSGeo California
blog.light42.com

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-devel/attachments/20190219/9ef82e57/attachment.html>


More information about the postgis-devel mailing list