Re: Write-heavy pg_stats_collector on mostly idle server - Mailing list pgsql-performance

From Josh Berkus
Subject Re: Write-heavy pg_stats_collector on mostly idle server
Date
Msg-id 4D509EE1.4070305@agliodbs.com
Whole thread Raw
In response to Re: Write-heavy pg_stats_collector on mostly idle server  (Cédric Villemain <cedric.villemain.debian@gmail.com>)
Responses Re: Write-heavy pg_stats_collector on mostly idle server  (Jakub Ouhrabka <kuba@comgate.cz>)
List pgsql-performance
> Ooops.
> It looks like you are right, see ./src/backend/postmaster/pgstat.c
>
> 3c2313f4 (Tom Lane           2008-11-03 01:17:08 +0000 2926)
>  if (last_statwrite < last_statrequest)
> 70d75697 (Magnus Hagander    2008-08-05 12:09:30 +0000 2927)
>          pgstat_write_statsfile(false);

This is a different issue.  This is happening because we have a bunch of
databases (25 to 35) and as a result autovacuum is requesting the stats
file rather frequently.  And autovacuum for whatever reason won't accept
a stats file more than 10ms old, so it pretty much rewrites the stats
file on every request.

At least, that's my reading of it after poking around and talking to
Gierth.

It seems like that 10ms window for autovac is way too small.

--
                                  -- Josh Berkus
                                     PostgreSQL Experts Inc.
                                     http://www.pgexperts.com

pgsql-performance by date:

Previous
From: Sylvain Rabot
Date:
Subject: Indexes with condition using immutable functions applied to column not used
Next
From: felix
Date:
Subject: Re: Really really slow select count(*)