From mboxrd@z Thu Jan  1 00:00:00 1970
From: Stephen Hemminger <shemminger@vyatta.com>
Subject: Re: [RFC] sched: CHOKe packet scheduler (v0.2)
Date: Thu, 6 Jan 2011 23:10:36 -0800
Message-ID: <20110106231036.0252641f@s6510>
References: <20110104162930.6fa672e3@nehalam>
	<1294208375.3420.46.camel@edumazet-laptop>
	<20110105091718.02f8a00f@nehalam>
	<1294248332.10633.25.camel@edumazet-laptop>
	<20110105112104.64ad3c86@nehalam>
	<1294286850.2723.65.camel@edumazet-laptop>
	<20110106205549.0de56de1@nehalam>
	<AANLkTi=C6c2md1_MB6dqvVHx2OL0Vj5Lgosi2C=kK9YG@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: Eric Dumazet <eric.dumazet@gmail.com>,
	David Miller <davem@davemloft.net>, netdev@vger.kernel.org
To: Changli Gao <xiaosuo@gmail.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail.vyatta.com ([76.74.103.46]:46673 "EHLO mail.vyatta.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1752541Ab1AGHKq convert rfc822-to-8bit (ORCPT
	<rfc822;netdev@vger.kernel.org>); Fri, 7 Jan 2011 02:10:46 -0500
In-Reply-To: <AANLkTi=C6c2md1_MB6dqvVHx2OL0Vj5Lgosi2C=kK9YG@mail.gmail.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Fri, 7 Jan 2011 13:39:26 +0800
Changli Gao <xiaosuo@gmail.com> wrote:

> On Fri, Jan 7, 2011 at 12:55 PM, Stephen Hemminger
> <shemminger@vyatta.com> wrote:
> > On Thu, 06 Jan 2011 05:07:30 +0100
> > Eric Dumazet <eric.dumazet@gmail.com> wrote:
> >
> >> Le mercredi 05 janvier 2011 =E0 11:21 -0800, Stephen Hemminger a =E9=
crit :
> >> > This implements the CHOKe packet scheduler based on the existing
> >> > Linux RED scheduler based on the algorithm described in the pape=
r.
> >> >
> >> > The core idea is:
> >> > =A0 For every packet arrival:
> >> > =A0 =A0 Calculate Qave
> >> > =A0 =A0 if (Qave < minth)
> >> > =A0 =A0 =A0 =A0 =A0Queue the new packet
> >> > =A0 =A0 else
> >> > =A0 =A0 =A0 =A0 =A0Select randomly a packet from the queue
> >> > =A0 =A0 =A0 =A0 =A0if (both packets from same flow)
> >> > =A0 =A0 =A0 =A0 =A0then Drop both the packets
> >> > =A0 =A0 =A0 =A0 =A0else if (Qave > maxth)
> >> > =A0 =A0 =A0 =A0 =A0 =A0 =A0 Drop packet
> >> > =A0 =A0 =A0 =A0 =A0else
> >> > =A0 =A0 =A0 =A0 =A0 =A0 =A0 Admit packet with probability p (sam=
e as RED)
> >> >
> >> > See also:
> >> > =A0 Rong Pan, Balaji Prabhakar, Konstantinos Psounis, "CHOKe: a =
stateless active
> >> > =A0 =A0queue management scheme for approximating fair bandwidth =
allocation",
> >> > =A0 Proceeding of INFOCOM'2000, March 2000.
> >> >
> >> > Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
> >> >
> >>
> >> To be really useful in a wide range of environments, I believe tha=
t :
> >>
> >> - CHOKe should be able to use an external flow classifier (like sa=
y...
> >> SFQ) to compute a token and compare two skbs by this token instead=
 of
> >> custom rxhash or whatever. (rxhash can be the default in absence o=
f flow
> >> classifier). Probably you need to store the token in skb->cb[] to =
avoid
> >> calling tc_classify() several times for a given packet.
> >>
> >> http://lwn.net/Articles/236200/
> >> http://kerneltrap.org/mailarchive/linux-netdev/2008/1/31/667679
> >
> > Probably should split SFQ flow hash stuff into core code for reuse.
> >
> >
>=20
> We need not do that, since we have sch_drr and cls_flow. :)

I prefer that the qdisc be useable without any explicit flow classifica=
tion.
I.e like SFQ it should fall back to a sensible flow matching. DRR and o=
thers
put everything in one flow, if no filters are used.