Re: Performance issue with Pointcloud extension

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Eric Lemoine <eric(dot)lemoine(at)oslandia(dot)com>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: Performance issue with Pointcloud extension
Date: 2017-06-08 18:07:13
Message-ID: 5413.1496945233@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Eric Lemoine <eric(dot)lemoine(at)oslandia(dot)com> writes:
> The initial "select pc_typmod_pcid(1)" query completely screws the
> connection.
> "select pc_typmod_pcid(1)" is just an example of a simple query that
> triggers the problem. There are many others. But it has to be a query
> using the Pointcloud extension.

My guess is that it's got nothing to do with the specific query, but
that Pointcloud is installing some hook functions when it's loaded,
and that one or another of those hooks is killing performance for
subsequent queries, either by taking too long in itself or by
defeating some important optimization. Hard to speculate further
with just this much data.

> I have no problem if I start with the main query (my "select points
> from" query of interest). And running the "select pc_typmod_pcid(1)"
> query in the middle does not cause any problem. It has to be run first
> on the connection to do the harm.

Have you experimented with other queries that don't involve PostGIS?
I'm wondering if your hook-installation code fails to work properly
unless PostGIS was loaded first. This would be easier to credit if
there are hooks both extensions try to get into.

regards, tom lane

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Zhu, Joshua 2017-06-08 18:47:48 Re: How does BDR replicate changes among nodes in a BDR group
Previous Message Jonathan Zacharuk 2017-06-08 17:45:54 Is there a way to verify a signed digest within Postgres?