Horton

Re: [HACKERS] kqueue

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Matteo Beccati <php@beccati.com>
Cc: Thomas Munro <thomas.munro@gmail.com>, Peter Eisentraut <peter.eisentraut@2ndquadrant.com>, Rui DeSousa <rui@crazybean.net>, Torsten Zuehlsdorff <mailinglists@toco-domains.de>, Andres Freund <andres@anarazel.de>, Heikki Linnakangas <hlinnaka@iki.fi>, Marko Tiikkaja <marko@joh.to>, Alvaro Herrera <alvherre@2ndquadrant.com>, Robert Haas <robertmhaas@gmail.com>, Pg Hackers <pgsql-hackers@postgresql.org>, Noah Misch <noah@leadboat.com>
Date: 2020-01-22T19:19:11Z
Lists: pgsql-hackers 
I wrote:
> This just says it doesn't lock up, of course.  I've not attempted
> any performance-oriented tests.

I've now done some light performance testing -- just stuff like
pgbench -S -M prepared -c 20 -j 20 -T 60 bench

I cannot see any improvement on either FreeBSD 12 or NetBSD 8.1,
either as to net TPS or as to CPU load.  If anything, the TPS
rate is a bit lower with the patch, though I'm not sure that
that effect is above the noise level.

It's certainly possible that to see any benefit you need stress
levels above what I can manage on the small box I've got these
OSes on.  Still, it'd be nice if a performance patch could show
some improved performance, before we take any portability risks
for it.

			regards, tom lane

Commits

  1. Add kqueue(2) support to the WaitEventSet API.

  2. Add WL_EXIT_ON_PM_DEATH pseudo-event.

  3. Be conservative about alignment requirements of struct epoll_event.

  4. Make idle backends exit if the postmaster dies.