From: Dan Kegel <dank@kegel.com>
To: Arjen Wolfs <arjen@euro.net>
Cc: coder-com@undernet.org, feedback@distributopia.com,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [Coder-Com] Re: PROBLEM: high system usage / poor SMPnetwork performance
Date: Sun, 03 Feb 2002 11:16:09 -0800 [thread overview]
Message-ID: <3C5D8C79.4C0792C5@kegel.com> (raw)
In-Reply-To: <3C56E327.69F8B70F@kegel.com> <001901c1a900$e2bc7420$0201010a@frodo> <3C58D50B.FD44524F@kegel.com> <001d01c1aa8e$2e067e60$0201010a@frodo> <5.1.0.14.2.20020203173247.02c946e8@pop.euronet.nl>
Arjen Wolfs wrote:
> The ircu version that supports kqueue and /dev/poll is currently being
> beta-tested on a few servers on the Undernet. The graph at
> http://www.break.net/ircu10-to-11.png shows the load average (multiplied by
> 100) on a on a server with 3000-4000 clients using poll(), and /dev/poll.
> The difference is obviously quite dramatic, and the same effect is being
> seen with kqueue. You could also try some of the /dev/poll patches for
> linux, which migth save you writing a new engine. Note that ircu 2.10.11 is
> still beta though, and is known to crash in mysterious ways from time to time.
None of the original /dev/poll patches for Linux were much
good, I seem to recall; they had scaling problems and bugs.
The /dev/epoll patch is good, but the interface is different enough
from /dev/poll that ircd would need a new engine_epoll.c anyway.
(It would look like a cross between engine_devpoll.c and engine_rtsig.c,
as it would need to be notified by os_linux.c of any EWOULDBLOCK return values.
Both rtsigs and /dev/epoll only provide 'I just became ready' notification,
but no 'I'm not ready anymore' notification.)
And then there's /dev/yapoll (http://www.distributopia.com), which
I haven't tried yet (I don't think the author ever published the patch?).
Anyway, the new engine wouldn't be too hard to write, and
would let irc run fast without a patched kernel.
- Dan
next prev parent reply other threads:[~2002-02-03 19:10 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-01-29 18:00 PROBLEM: high system usage / poor SMP network performance Dan Kegel
2002-01-29 20:09 ` Vincent Sweeney
2002-01-31 5:24 ` Dan Kegel
[not found] ` <001d01c1aa8e$2e067e60$0201010a@frodo>
2002-02-03 8:03 ` Dan Kegel
2002-02-03 8:36 ` Andrew Morton
2002-02-04 14:57 ` [Coder-Com] " Darren Smith
2002-02-04 17:41 ` Aaron Sethman
2002-02-04 18:11 ` Darren Smith
2002-02-04 18:30 ` Aaron Sethman
2002-02-04 18:48 ` Kev
2002-02-04 18:59 ` Aaron Sethman
2002-02-04 18:53 ` Doug McNaught
2002-02-08 22:11 ` James Antill
2002-02-12 18:48 ` Vincent Sweeney
2002-02-03 19:22 ` Kev
[not found] ` <5.1.0.14.2.20020203173247.02c946e8@pop.euronet.nl>
2002-02-03 19:16 ` Dan Kegel [this message]
2002-02-04 0:07 ` [Coder-Com] Re: PROBLEM: high system usage / poor SMPnetwork performance Kev
2002-02-04 0:37 ` Dan Kegel
2002-02-04 0:59 ` Aaron Sethman
2002-02-04 1:16 ` Dan Kegel
2002-02-04 1:30 ` Aaron Sethman
2002-02-04 1:38 ` Dan Kegel
2002-02-04 4:38 ` Aaron Sethman
2002-02-04 5:35 ` Dan Kegel
2002-02-04 5:43 ` Aaron Sethman
2002-02-04 6:11 ` Daniel Phillips
2002-02-04 6:26 ` Aaron Sethman
2002-02-04 6:29 ` Daniel Phillips
2002-02-04 6:39 ` Aaron Sethman
2002-02-04 2:55 ` Kev
2002-02-04 3:25 ` Dan Kegel
2002-02-04 4:47 ` Aaron Sethman
2002-02-04 5:10 ` Kev
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=3C5D8C79.4C0792C5@kegel.com \
--to=dank@kegel.com \
--cc=arjen@euro.net \
--cc=coder-com@undernet.org \
--cc=feedback@distributopia.com \
--cc=linux-kernel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox