From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-vk0-x234.google.com (mail-vk0-x234.google.com [IPv6:2607:f8b0:400c:c05::234]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 11F3C3B25E for ; Tue, 11 Oct 2016 08:09:32 -0400 (EDT) Received: by mail-vk0-x234.google.com with SMTP id 192so16147061vkl.2 for ; Tue, 11 Oct 2016 05:09:32 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bitamins-net.20150623.gappssmtp.com; s=20150623; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc; bh=FhXg7a+nLIUml46DLs0v/2fY0gu2QyH2RuA4KSoe6rQ=; b=O2uJgo9+Fb4ITLwt+oOByKjHuuRlGPbLnzcFQvtkximByn6dPLF0VjuCoECFU1mERc 00IjXewUTIYD1AqZoJDCB9LcVKwGd8VeV/1tddLs+ZNogcrudMX/al9YkoI3rDeuw8EB wR8Y2pqyxoOj6kPAKayiG3GFSRp0EbwOt0qBYyNN3Qf6JxcdaXiXBOx783MCqhE7Rs/m NZk1pJ73qawmS6ehing2VRZkIikyjxl131aHcNF/z+y4ZyrL964vpHjcJGOL3BK1hZgY 34elR7uMSvJ2gKA0EET91+0LhKK4D0w+YDJPHypmjySLego22pLgvuSbiZpybjf6FWY9 SUpw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc; bh=FhXg7a+nLIUml46DLs0v/2fY0gu2QyH2RuA4KSoe6rQ=; b=fFuMGC92jEOfZ+9OpiRQo8NHo4wg778bRG4Xi5j7f16gHRV4qQygMdHWgvOQ79RPLd g6/EmDdeAe1plWOrnf3/2fXBXIyEhElNJ6dkuY5RWf5e3RK3TpmGbD5gtUdhhBihWyxy bk0bU0MRdADT+MIA32QFhQY/MZdRwXKgbNSBk28OYeb9KMp5Jm81aAlsn+ihU3CDzCE9 eR8hPdCpcxete/mB22Un1j1e/f5H2G88vIKQc8mmfYKEDB0VEpoPqecFT/GV36gMmZoZ tXA9Zf2t0zF3CNHLoIMUagFd4O1D1Lq3mgDPJ/BoQshaJFO51FdhfEVvowCf0LPm44RQ bXjA== X-Gm-Message-State: AA6/9Rmbn8yS8mMdTTDRIcRzqpU3R4dpyKvkIyMI8rTc84x6xtURhoOVQGsnYtmZn6afgcLo8b02ZmkMi6GbPA== X-Received: by 10.31.226.67 with SMTP id z64mr2770901vkg.141.1476187772603; Tue, 11 Oct 2016 05:09:32 -0700 (PDT) MIME-Version: 1.0 Received: by 10.159.35.12 with HTTP; Tue, 11 Oct 2016 05:09:32 -0700 (PDT) X-Originating-IP: [190.62.179.253] In-Reply-To: References: <264e5761-c70b-3518-0804-dea1e92bec08@darbyshire-bryant.me.uk> <4C68E9EC-FABD-462B-A387-69B53E96A540@gmail.com> From: "Luis E. Garcia" Date: Tue, 11 Oct 2016 06:09:32 -0600 Message-ID: To: Jonathan Morton Cc: Kevin Darbyshire-Bryant , "cake@lists.bufferbloat.net" Content-Type: multipart/alternative; boundary=001a114df32ccee5a4053e95bf65 Subject: Re: [Cake] Master branch updated 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: Tue, 11 Oct 2016 12:09:33 -0000 --001a114df32ccee5a4053e95bf65 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable If we can replicate the results of your test then I would say we're onto something. On Monday, 10 October 2016, Jonathan Morton wrote: > > > On 4 Oct, 2016, at 19:28, Jonathan Morton > wrote: > > > >> Ha ha! I don't know if you're back from shopping yet...and I'm not > sure that I've broken it (cobalt branch)...but it has broken my router! > > > > Hmm. It=E2=80=99s been running all day with plenty of traffic over her= e - but > it did crash the very first time I loaded it, just not the second. I wil= l > need to exercise it some more, preferably on a non-critical machine. > > Okay, that bug is fixed and I=E2=80=99ve made further improvements to the > triple-isolate algorithm. It no longer needs quite as much spaghetti log= ic > in the fast path, and might even be easier to understand from reading the > code, since it=E2=80=99s now more obviously a modification of DRR++ rathe= r than a > brute-force wrapper around it. It should certainly give smoother behavio= ur > and be less CPU intensive in common cases. > > In brief, what I now do is to scale the *flow* quantum down by the higher > of the two hosts=E2=80=99 flow counts. I=E2=80=99ve even dealt with unde= rflow of the > quotient using a dithering mechanism, which should also ensure that flows > random-walk out of lockstep with each other. > > It works sufficiently well that I was able to set Cake to 2.5Mbit > besteffort triple-isolate, then watch a 720p YouTube video on one machine > while another was downloading a game update using a 30-flow swarm. I=E2= =80=99d > call that a success. > > Hammer away at it, and then we=E2=80=99ll see if we can merge it up to ma= ster. > > - Jonathan Morton > > _______________________________________________ > Cake mailing list > Cake@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/cake > --001a114df32ccee5a4053e95bf65 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable If we can replicate=C2=A0the results of your test then I would say we'r= e onto something.

On Monday, 10 October 2016, Jonathan = Morton <chromatix99@gmail.com> wrote:

> On 4 Oct, 2016, at 19:28, Jonathan Morton <
chr= omatix99@gmail.com> wrote:
>
>> Ha ha!=C2=A0 I don't know if you're back from shopping yet= ...and I'm not sure that I've broken it (cobalt branch)...but it ha= s broken my router!
>
> Hmm.=C2=A0 It=E2=80=99s been running all day with plenty of traffic ov= er here - but it did crash the very first time I loaded it, just not the se= cond.=C2=A0 I will need to exercise it some more, preferably on a non-criti= cal machine.

Okay, that bug is fixed and I=E2=80=99ve made further improvements to the t= riple-isolate algorithm.=C2=A0 It no longer needs quite as much spaghetti l= ogic in the fast path, and might even be easier to understand from reading = the code, since it=E2=80=99s now more obviously a modification of DRR++ rat= her than a brute-force wrapper around it.=C2=A0 It should certainly give sm= oother behaviour and be less CPU intensive in common cases.

In brief, what I now do is to scale the *flow* quantum down by the higher o= f the two hosts=E2=80=99 flow counts.=C2=A0 I=E2=80=99ve even dealt with un= derflow of the quotient using a dithering mechanism, which should also ensu= re that flows random-walk out of lockstep with each other.

It works sufficiently well that I was able to set Cake to 2.5Mbit besteffor= t triple-isolate, then watch a 720p YouTube video on one machine while anot= her was downloading a game update using a 30-flow swarm.=C2=A0 I=E2=80=99d = call that a success.

Hammer away at it, and then we=E2=80=99ll see if we can merge it up to mast= er.

=C2=A0- Jonathan Morton

_______________________________________________
Cake mailing list
Cake@lists.bufferbloat.net
h= ttps://lists.bufferbloat.net/listinfo/cake
--001a114df32ccee5a4053e95bf65--