From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH] ipv4: add DiffServ priority based routing Date: Fri, 15 Jan 2010 09:24:36 +0100 Message-ID: <4B502644.4070000@gmail.com> References: <20100112.121607.39835310.davem@davemloft.net> <201001141250.27287.schmto@hrz.tu-chemnitz.de> <4B4F132B.7020001@gmail.com> <20100114.165142.163322281.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: torsten.schmidt@s2006.tu-chemnitz.de, netdev@vger.kernel.org To: David Miller Return-path: Received: from gw1.cosmosbay.com ([212.99.114.194]:55637 "EHLO gw1.cosmosbay.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754874Ab0AOIYl (ORCPT ); Fri, 15 Jan 2010 03:24:41 -0500 In-Reply-To: <20100114.165142.163322281.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: Le 15/01/2010 01:51, David Miller a =E9crit : > From: Eric Dumazet > Date: Thu, 14 Jan 2010 13:50:51 +0100 >=20 >> Problem is, you might need many sets of counters... >=20 > Why nobody has suggested using existing kernel facilities for this is > beyond me. >=20 > We have a diffserv packet scheduler queueing discipline, so you can > classify traffic arbitrarily based upon the diffserv bits in the > packet and then shape them using that classification into different > packet scheduler queues. >=20 > Simply attach those child queues to that default pfifo qdisc, nothing > fancy. >=20 > Then you can dump the queue stats using 'tc'. Well, its a good idea but it has drawbacks, for example if your eth0 de= vice has 64 tx queues. Parsing tc output can be... interesting :)