I have multiple large polygon datasets stored in PostgreSQL 12 and will publish them using GeoServer. I want to try to use geohash to possibly speed up the performance of serving the polygon layers. I made a geohash index on one of the larger polygon datasets in PostgreSQL and it seems to have slightly improved performance but that just might be a hopeful placebo effect.

Anyone know of any way to use geohash to speed up serving large polygon data?

  • 2
    Query is usually a tiny fraction of the time spent serving maps. I/O is the longest. So using the optization creed ("optimize the slowest part first") you should probably be looking at vertex simplification first, as a way to make the PG rowspace smaller, and render or transmission time shorter. You left out details on how your polygons are large (rowcount? avg vertex count? area of envelopes?) and the organization of the features in the table (sorted on time? sorted on ID? sorted by latitude within UTM zone?) so it's difficult to recommend how to proceed.. – Vince Nov 14 '19 at 14:06
  • Usually the best way to speed up postgis queries is to make sure that postgresql is configured correctly – Ian Turton Nov 15 '19 at 10:34

Browse other questions tagged or ask your own question.