Poor performance PostgreSQL13/PostGIS 3.x

From: "Lugosi, Jim" <JimLug(at)clackamas(dot)us>
To: "pgsql-hackers(at)lists(dot)postgresql(dot)org" <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Poor performance PostgreSQL13/PostGIS 3.x
Date: 2022-01-20 22:31:15
Message-ID: 792d4b48d5b2404cb1a521b25536eb49@clackamas.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi folks,

We are struggling to figure out what is going on. We are migrating from PostgreSQL 9.6 to PostgreSQL 13 w/ PostGIS. Our 9.6 version was compiled from source and the new version (13) was installed using Yum. BTW, the new version is on a VM that has 16GB of memory, two cores, and 500 GB of disk. In addition, we are using MapServer as our mapping engine and OpenLayers as the client side interface. Once we switch over to the new version of PostgreSQL, the performance takes a big nose dive. We have being tweaking and tuning the database and it appears to be happy but the response times from mapfile requests are 3 -7 seconds. Previously, the response time was below a second.

Another point is that we populated the new database from the old (9.6), using pg_dump. Could this be causing issues? Should we load the data from scratch? We use ogr2ogr (GDAL) to help assist with loading of spatial data. Anyway, not really sure what the problem is.

Lastly, why am I seeing so many requests as to the PostGIS version. It appears that every map request sends the following query "SELECT PostGIS_Version();", which in turn takes up a connection.

Any suggestions would be greatly appreciated?

__:)
_ \<,_
(*)/ (*)
James Lugosi
Clackamas County GISP
IS Software Specialist, Senior
121 Library Court, Oregon City OR 97045
503-723-4829

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Justin Pryzby 2022-01-20 22:43:52 Re: Poor performance PostgreSQL13/PostGIS 3.x
Previous Message Peter Geoghegan 2022-01-20 22:00:12 Re: Removing more vacuumlazy.c special cases, relfrozenxid optimizations