[Top] [All Lists]

Re: _High_ CPU usage while routing (mostly) small UDP packets

To: Karsten Desler <kdesler@xxxxxxxxxx>
Subject: Re: _High_ CPU usage while routing (mostly) small UDP packets
From: jamal <hadi@xxxxxxxxxx>
Date: 07 Dec 2004 23:31:58 -0500
Cc: netdev@xxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, "David S. Miller" <davem@xxxxxxxxxxxxx>, Robert Olsson <Robert.Olsson@xxxxxxxxxxx>, P@xxxxxxxxxxxxxx
In-reply-to: <20041207211035.GA20286@xxxxxxxxxxxxxxxxxxxx>
Organization: jamalopolous
References: <20041206205305.GA11970@xxxxxxxxxx> <20041207211035.GA20286@xxxxxxxxxxxxxxxxxxxx>
Reply-to: hadi@xxxxxxxxxx
Sender: netdev-bounce@xxxxxxxxxxx
On Tue, 2004-12-07 at 16:10, Karsten Desler wrote:
> Karsten Desler <kdesler@xxxxxxxxxx> wrote:
> > Current packetload on eth0 (and reversed on eth1):
> >   115kpps tx
> >   135kpps rx
> I totally forgot to mention: There are approximately 100k concurrent
> flows.

;-> Aha. That would make a huge difference. I know of noone
who has actually done this level of testing. I have tried upto about 50K
flows myself in early 2.6.x and was eventually compute bound.
Really try compiling out totaly iptables/netfilter - it will make a
You may also want to try something like LC trie algorithm that Robert
and co are playing with to see if it makes a difference with this many


<Prev in Thread] Current Thread [Next in Thread>