| From: | Andres Freund <andres(at)anarazel(dot)de> |
|---|---|
| To: | Justin Pryzby <pryzby(at)telsasoft(dot)com> |
| Cc: | Laurenz Albe <laurenz(dot)albe(at)cybertec(dot)at>, Magnus Hagander <magnus(at)hagander(dot)net>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: pgstat_send_connstats() introduces unnecessary timestamp and UDP overhead |
| Date: | 2021-08-25 08:17:19 |
| Message-ID: | 20210825081719.ppagimzbpjce75ts@alap3.anarazel.de |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi,
On 2021-08-20 14:27:20 -0500, Justin Pryzby wrote:
> On Tue, Aug 17, 2021 at 02:14:20AM -0700, Andres Freund wrote:
> > Doubling the number of UDP messages in common workloads seems also problematic
> > enough that it should be addressed for 14. It increases the likelihood of
> > dropping stats messages on busy systems, which can have downstream impacts.
>
> I think by "common workloads" you mean one with many, shortlived sessions.
You don't need short-lived sessions. You just need sessions that don't
process queries all the time (so that there's only one or a few queries
within each PGSTAT_STAT_INTERVAL). The connection stats aren't sent once
per session, they're sent once per PGSTAT_STAT_INTERVAL.
Greetings,
Andres Freund
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2021-08-25 08:20:03 | Re: pgstat_send_connstats() introduces unnecessary timestamp and UDP overhead |
| Previous Message | Masahiko Sawada | 2021-08-25 07:48:59 | Re: Added schema level support for publication. |