From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mout.gmx.net (mout.gmx.net [212.227.15.18]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id A17743B25E for ; Wed, 1 Jun 2016 11:19:12 -0400 (EDT) Received: from [172.17.3.79] ([134.76.241.253]) by mail.gmx.com (mrgmx001) with ESMTPSA (Nemesis) id 0MJByE-1b5N243bli-002rBl; Wed, 01 Jun 2016 17:19:10 +0200 Content-Type: text/plain; charset=utf-8 Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\)) From: moeller0 In-Reply-To: <54AE2960-D312-479D-8411-96993F087EF7@gmail.com> Date: Wed, 1 Jun 2016 17:19:09 +0200 Cc: Benjamin Cronce , cake@lists.bufferbloat.net Content-Transfer-Encoding: quoted-printable Message-Id: References: <574EB29B.1000405@darbyshire-bryant.me.uk> <574EB550.5020005@taht.net> <574EB6E2.2020006@darbyshire-bryant.me.uk> <4DDB6EED-A66B-4E34-B233-8DC55F663EBD@gmx.de> <87shwxb1fk.fsf@toke.dk> <574ECB5E.7090605@darbyshire-bryant.me.uk> <0026A232-9D17-40FD-83A4-8575C6FFB8C3@gmx.de> <54AE2960-D312-479D-8411-96993F087EF7@gmail.com> To: Jonathan Morton X-Mailer: Apple Mail (2.2104) X-Provags-ID: V03:K0:fTVe/Bef4JES+iKZug5f7T1l6iX/y+qZF8m4ajT0fGQjjOTx1hZ I/1Ui2ZU2EltBydvXKqSeKkbkcn7anReC3LUUQIj5Q+2kg8a9bg9WjWlJQHyrrSafCz9p8E hjdHWI1oSUKwDPzf++91heW6no91G4cGtKYayseJ4js2cO75qdhzSFp94OipXrQQs49SJu6 LUE3PRfyyg29QyY8Iiskw== X-UI-Out-Filterresults: notjunk:1;V01:K0:m31UvffNK4s=:xNc0qpaIQ2a6sTQdQxyG9C UjyGKlV3NqtYs4MVF/bKkr6oqkYkn2Ua3gI5R+BghcfP4BpdJghlwmoowSygfyczXnSKIO2Hm qM1lBK38UPLyQP+cgf7eGxbuUCMRfCaivoxlU7YpTMOKK8ztzX4tGGkzpVTGXaNYJujbJrRdp Hud0wYeC2wTINCvVCoGcYni8gnqvqOrkb5XwBE0wFZTuS41hjd6yzGMtWn7Xi0T828sNBNarQ 9SeIzAOOnDdqVYP+OW2mSHyLGtgCaHlGBAo4YNbrTGwMv3cg9c6J0Xa86aWQcKTrtEcdBlDx6 1ZnXcl0J3NzWmnX4VugAIcCKATTL/jdW0WGAV4Alv8E1/qlykZLWiWOAdE5PAUH6rNBBH44Pm emsd+Y8MKF9Y/lx492aRnMkzJc2ZcJ527dMGqFfDPMzAp25T4pFeiHxDMbhpgF6kydSDbEj7K HnXPqkOw3oLbCVeCa7er/77GdbbncuWy3WRXjVfF5dfrIGnZtTn7dtxS3N6nGwaXCtxRzixRj f0c15K88tGRJ0yikJynYxg/9cNI7LPQPwmHBydx5jmuVB5M9JByoZH50nZH7BXoXE5RJuI2VU oDZyrQYWW1gt1U+xST3DfLs5z5IwdAz9NrD127VwQfwkm6p+ksZU3xtp/5EIXVK1ZIqPwV3+6 3RMktmRLQ0Go72IBvbH1Y3ZQdDPmawYM89bJZowKPmRnBB/2FFfH9Jm0OTlzXO8ajNs8abMdJ etzrS9KdB/pGwdUF5pIN9SH2p6iylwJGD0Tt7hxVnBWoHTSzHMNRQw8cMSxe/GXRQkAjLlVSs h6/nzKg Subject: Re: [Cake] [lede-project/source] Add support for cake qdisc (#72) X-BeenThere: cake@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: Cake - FQ_codel the next generation List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Jun 2016 15:19:13 -0000 Hi Jonathan, > On Jun 1, 2016, at 15:51 , Jonathan Morton = wrote: >=20 >=20 >> On 1 Jun, 2016, at 15:25, Benjamin Cronce wrote: >>=20 >> 1) Ideally, regardless of platform, should an AQM or scheduler have = the responsibility of changing anything other than ECN? >=20 > This was in part my original objection to having the squash/wash = feature in Cake. >=20 > The other part is that if we are going to rewrite the rest of the TOS = byte (not just the ECN bits), then we should do it properly, which = requires a rather substantial extension to the configuration API, even = if we only try to cover the most obvious use-cases. To play devils advocate here, re-mapping to zero might be a = special case, but probably the single most important special case, so = lack of generality is not the strongest argument against it (as long as = the goal is =E2=80=9Cgood enough=E2=80=9D). But I already conceded, it = is contentious so needs to go, cake is your =E2=80=9Cbaby=E2=80=9D so = you have the honor to make such calls=E2=80=A6 >=20 > This would then be a =E2=80=9Csemi proprietary=E2=80=9D DSCP = configurator, acting independently of Cake=E2=80=99s core AQM and = shaping functions, which would have to be duplicated in other AQMs which = had similar aims to Cake. That=E2=80=99s not a good thing, and feeds = back into the first point. Not necessarily, one could also argue that the first = implementation is free to just =E2=80=9Cdo it=E2=80=9D and only after it = proofed itself useful one needs to think about generalizing it. >=20 > Hence the correct solution is to put DSCP rewriting elsewhere, making = it reusable. >=20 > In Linux, doing ingress DSCP rewriting before it hits the ingress = qdisc presently requires the rewriter itself to be a qdisc, but this can = have Cake as a child qdisc. For the simple =E2=80=9Cclearing=E2=80=9D = case iptables can be used instead, as long as Cake is configured to = ignore the inbound DSCP using the =E2=80=9Cbesteffort=E2=80=9D flag. I seem to recall that iptables/nftables folks want to be able to = =E2=80=9Cconnect=E2=80=9D to the ingress ssytem, so maybe that solves = the whole issue (short of making the common thing (re-map to zero) = simple) All that said, the decision is made and cake is on its path to upstream = somehow, whoooot. Best Regards Sebastian >=20 > - Jonathan Morton >=20