From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from vps.slashdirt.org (vps.slashdirt.org [144.91.108.218]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id E34843B29D for ; Wed, 22 Apr 2020 04:28:38 -0400 (EDT) Received: from chuck.tardis.lan (tardis.herebedragons.eu [171.22.3.161]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by vps.slashdirt.org (Postfix) with ESMTPSA id 9C9C260075; Wed, 22 Apr 2020 10:28:37 +0200 (CEST) DKIM-Filter: OpenDKIM Filter v2.11.0 vps.slashdirt.org 9C9C260075 DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=slashdirt.org; s=mail; t=1587544117; bh=rYf2pw1855Lfah/xdttTi6vc0BvjwkLD6t99IL8MMm8=; h=From:Subject:Date:In-Reply-To:Cc:To:References:From; b=nA0MANZ//yJ4QOUlpMfwTkgAWm6xAo9ZoAvtHJWibNNC5uTGbnP0DZQbxWJfMkQ6q 64TzhnmeFudrEudb8Zwph1YqDkwV9kP8EzJVMqM35K/3eeOj7GSX14SQ8ikoRQQhEJ j3miphQiPgJBZ/AsjhjJDRkct49X88bnC5cxY6O0= From: Thibaut Message-Id: Content-Type: multipart/alternative; boundary="Apple-Mail=_2844F9D9-1C1E-423E-B29D-94BD946A0212" Mime-Version: 1.0 (Mac OS X Mail 12.4 \(3445.104.14\)) Date: Wed, 22 Apr 2020 10:28:36 +0200 In-Reply-To: Cc: Jonathan Morton , Cake List To: Dave Taht References: <75FEC2D9-BFC8-4FA2-A972-D11A823C5528@gmail.com> <603DFF79-D0C0-41BD-A2FB-E40B95A9CBB0@gmail.com> X-Mailer: Apple Mail (2.3445.104.14) Subject: Re: [Cake] Advantages to tightly tuning latency 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, 22 Apr 2020 08:28:39 -0000 --Apple-Mail=_2844F9D9-1C1E-423E-B29D-94BD946A0212 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 > Le 22 avr. 2020 =C3=A0 00:44, Dave Taht a =C3=A9cr= it : >=20 > On Tue, Apr 21, 2020 at 3:33 PM Jonathan Morton = wrote: >>=20 >>> On 22 Apr, 2020, at 1:25 am, Thibaut wrote: >>>=20 >>> My curiosity is piqued. Can you elaborate on this? What does free.fr = do? >>=20 >> They're a large French ISP. They made their own CPE devices, and = debloated both them and their network quite a while ago. In that sense, = at least, they're a model for others to follow - but few have. >>=20 >> - Jonathan Morton >=20 > they are one of the few ISPs that insisted on getting full source code > to their DSL stack, and retained the chops to be able to modify it. I > really admire their revolution v6 product. First introduced in 2010, > it's been continuously updated, did ipv6 at the outset, got fq_codel > when it first came out, and they update the kernel regularly. All > kinds of great features on it, and ecn is enabled by default for those > also (things like samba). over 3 million boxes now I hear.... >=20 > with <1ms of delay in the dsl driver, they don't need to shape, they > just run at line rate using three tiers of DRR that look a lot like > cake. They shared their config with me, and before I lost heart for > future internet drafts, I'd stuck it here: >=20 > = https://github.com/dtaht/bufferbloat-rfcs/blob/master/home_gateway_queue_m= anagement/middle.mkd = Very interesting, thanks. I wonder if they trickled down these = improvements to the older V5: I had to plug my old V5 back after my = DSLAM was moved to native IPV6 (I couldn=E2=80=99t find how to talk to = it over VDSL2 modem, as I did before the switch), and though there was a = massive drop in uplink bandwidth (from 10Mbps VDSL2 to 1Mbps ADSL), I = noticed that I no longer needed cake on the router wan interface. = Latency remained very well controlled without having to do anything = special. In fact, enabling cake with the previous settings was wrecking = havoc! (Maybe it interfered with whatever Free is doing in the box). > Occasionally they share some data with me. Sometimes I wish I lived in > paris just so I could have good internet! (their fiber offering is > reasonably buffered (not fq_codeled) and the wifi... maybe I can get > them to talk about what they did) You don=E2=80=99t have to live in Paris to enjoy good internet: I=E2=80=99= m currently stranded in the countryside and I enjoy a better connection = than many a Parisian, thanks to a public/private FTTH network that = appears to be very well handled by my current ISP (K-Net) :) > When free.fr shipped fq_codel 2 months after we finalized it, I > figured the rest of the world was only months behind. How hard is it > to add 50 lines of BQL oriented code to a DSL firmware? Heh. Cheers, Thibaut= --Apple-Mail=_2844F9D9-1C1E-423E-B29D-94BD946A0212 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8

Le 22 avr. 2020 =C3=A0 00:44, Dave Taht <dave.taht@gmail.com>= a =C3=A9crit :

On Tue, Apr 21, 2020 at 3:33 PM Jonathan = Morton <chromatix99@gmail.com> wrote:

On 22 Apr, 2020, at 1:25 am, Thibaut <hacks@slashdirt.org>= wrote:

My curiosity is piqued. Can you = elaborate on this? What does free.fr do?

They're a large French ISP.  They made their own CPE = devices, and debloated both them and their network quite a while ago. =  In that sense, at least, they're a model for others to follow - = but few have.

- Jonathan Morton

they are one of the few ISPs that = insisted on getting full source code
to their DSL stack, = and retained the chops to be able to modify it. I
really = admire their revolution v6 product. First introduced in 2010,
it's been continuously updated, did ipv6 at the outset, got = fq_codel
when it first came out, and they update the = kernel regularly. All
kinds of great features on it, and = ecn is enabled by default for those
also (things like = samba). over 3 million boxes now I hear....

with <1ms of delay in the dsl driver, they don't need to = shape, they
just run at line rate using three tiers of DRR = that look a lot like
cake. They shared their config with = me, and before I lost heart for
future internet drafts, = I'd stuck it here:

https://github.com/dtaht/bufferbloat-rfcs/blob/master/home_gate= way_queue_management/middle.mkd

Very = interesting, thanks. I wonder if they trickled down these improvements = to the older V5: I had to plug my old V5 back after my DSLAM was moved = to native IPV6 (I couldn=E2=80=99t find how to talk to it over VDSL2 = modem, as I did before the switch), and though there was a massive drop = in uplink bandwidth (from 10Mbps VDSL2 to 1Mbps ADSL), I noticed that I = no longer needed cake on the router wan interface. Latency remained very = well controlled without having to do anything special. In fact, enabling = cake with the previous settings was wrecking havoc! (Maybe it interfered = with whatever Free is doing in the box).

Occasionally they share some data with me. Sometimes I wish I = lived in
paris just so I could have good internet! (their = fiber offering is
reasonably buffered (not fq_codeled) and = the wifi... maybe I can get
them to talk about what they = did)

You = don=E2=80=99t have to live in Paris to enjoy good internet: I=E2=80=99m = currently stranded in the countryside and I enjoy a better connection = than many a Parisian, thanks to a public/private FTTH network that = appears to be very well handled by my current ISP (K-Net) = :)

When free.fr shipped fq_codel 2 months after we finalized it, = I
figured the rest of the world was only months behind. = How hard is it
to add 50 lines of BQL oriented code to a = DSL firmware?

Heh.

Cheers,
Thibaut
= --Apple-Mail=_2844F9D9-1C1E-423E-B29D-94BD946A0212--