Horton

Re: pgstat_send_connstats() introduces unnecessary timestamp and UDP overhead

Laurenz Albe <laurenz.albe@cybertec.at>

View thread View on postgr.es
From: Laurenz Albe <laurenz.albe@cybertec.at>
To: Andres Freund <andres@anarazel.de>
Cc: magnus@hagander.net, Michael Paquier <michael@paquier.xyz>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, pgsql-hackers@postgresql.org
Date: 2021-09-06T07:37:55Z
Lists: pgsql-hackers 
On Mon, 2021-09-06 at 00:22 -0700, Andres Freund wrote:
> On 2021-09-06 09:12:58 +0200, Laurenz Albe wrote:
> > Reading your patch, I am still confused about the following:
> > There are potentially several calls to "pgstat_send_tabstat" in "pgstat_report_stat".
> > It seems to me that if it were called more than once, session statistics would
> > be reported and counted several times, which would be wrong.
> 
> The accumulator variables are set to zero after being sent. So no data should
> be sent multiple times as far as I can tell.

Ah, I missed that.  Then I see no problem with that approach.

Yours,
Laurenz Albe

Commits

  1. Fix performance regression from session statistics.

  2. Add pg_stat_database counters for sessions and session time