From: P@draigBrady.com
To: Karsten Desler <kdesler@soohrt.org>
Cc: "David S. Miller" <davem@davemloft.net>,
netdev@oss.sgi.com, linux-kernel@vger.kernel.org
Subject: Re: _High_ CPU usage while routing (mostly) small UDP packets
Date: Tue, 07 Dec 2004 10:47:52 +0000 [thread overview]
Message-ID: <41B58A58.8010007@draigBrady.com> (raw)
In-Reply-To: <20041206224107.GA8529@soohrt.org>
[-- Attachment #1: Type: text/plain, Size: 1149 bytes --]
Karsten Desler wrote:
> * David S. Miller wrote:
>
>>It's spending nearly half of it's time in iptables.
>>Try to consolidate your rules if possible. This is the
>>part of netfilter that really doesn't scale well at all.
>>
>
> Removing the iptables rules helps reducing the load a little, but the
> majority of time is still spent somewhere else.
Well with NAPI it can be hard to tell CPU usage.
You may need to use something like cyclesoak to get
a true idea of CPU left.
Also have a look at http://www.hipac.org/ as netfilter
has silly scalability properties.
I also notice that a lot of time is spent allocating
and freeing the packet buffers (and possible hidden
time due to cache misses due to allocating on one
CPU and freeing on another?).
How many [RT]xDescriptors do you have configured by the way?
Anyway attached is a small patch that I used to make the e1000
"own" the packet buffers, and hence it does not alloc/free
per packet at all. Now this has only been tested in one
configuration where I was just sniffing the packets, so
definitely YMMV.
--
Pádraig Brady - http://www.pixelbeat.org
--
[-- Attachment #2: linux-2.4.20-5.2.52-realloc.diff --]
[-- Type: application/x-texinfo, Size: 7376 bytes --]
next prev parent reply other threads:[~2004-12-07 10:47 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-12-06 20:53 _High_ CPU usage while routing (mostly) small UDP packets Karsten Desler
2004-12-06 21:48 ` David S. Miller
2004-12-06 22:41 ` Karsten Desler
2004-12-06 23:56 ` Con Kolivas
2004-12-07 0:18 ` Karsten Desler
[not found] ` <E1CbSf8-00047p-00@calista.eckenfels.6bone.ka-ip.net>
2004-12-07 0:20 ` Karsten Desler
2004-12-07 2:46 ` jamal
2004-12-07 2:54 ` Karsten Desler
2004-12-07 3:18 ` jamal
2004-12-07 3:24 ` Karsten Desler
2004-12-07 3:30 ` jamal
2004-12-07 4:02 ` Karsten Desler
2004-12-07 10:21 ` Karsten Desler
2004-12-07 12:34 ` jamal
2004-12-07 13:14 ` Karsten Desler
2004-12-07 10:47 ` P [this message]
2004-12-07 11:21 ` Karsten Desler
2004-12-07 12:38 ` Robert Olsson
2004-12-07 12:50 ` Karsten Desler
2004-12-07 13:04 ` jamal
2004-12-07 13:11 ` Karsten Desler
2004-12-07 13:39 ` P
2004-12-07 18:38 ` Karsten Desler
2004-12-08 5:39 ` Willy Tarreau
2004-12-08 13:08 ` Karsten Desler
2004-12-08 13:27 ` jamal
2004-12-07 21:10 ` Karsten Desler
2004-12-07 22:40 ` Robert Olsson
2004-12-08 22:06 ` Karsten Desler
2004-12-08 4:31 ` jamal
2004-12-08 13:26 ` Karsten Desler
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=41B58A58.8010007@draigBrady.com \
--to=p@draigbrady.com \
--cc=davem@davemloft.net \
--cc=kdesler@soohrt.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@oss.sgi.com \
/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;
as well as URLs for NNTP newsgroup(s).