From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-it1-x12e.google.com (mail-it1-x12e.google.com [IPv6:2607:f8b0:4864:20::12e]) (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 3A4683B29E for ; Mon, 4 Feb 2019 04:18:23 -0500 (EST) Received: by mail-it1-x12e.google.com with SMTP id p197so19500397itp.0 for ; Mon, 04 Feb 2019 01:18:23 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=El4YwLabiP3THy2JW4w0y6qRjHW3P1/fe9ZKX9Fc9+M=; b=lxjN3auwpyZQjLQJeiJzU1QEflsSyDVdUpSexitIA1u1T9PYBxaEWu5T718TQ+doOk nM+34vJkqKAYNsY3tESSYLHkmUYha29wER9OWRaX3GPd1ZAvEmRhtKqfnfmels/AtrHj bqeOqpAzGkdH5hTG/O3qGQvmHk4is63RLHQKrPPnaQ/41B88ZVsdax2AO2QUlf7HjD9Z o14v+p+pkppVE2kB6VFyRF74fLl3vutqgcgrAPj9M14glrER1ue7X2AVernq/BmckCK+ vrQOpgxz3y61SM8pSITAZQtpYs/Tkf4dLH3A0UqF4LiU2bnpAXKIM+TVXjjhNko1MqPs ro2g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=El4YwLabiP3THy2JW4w0y6qRjHW3P1/fe9ZKX9Fc9+M=; b=WqpRpPb3o9YPTm9TRMx9qZl90U/qI/IkUY3b0Zz62DORXDjKQ0XqWQjRIOIkjeyqAQ e+YNOFUgJjbUaympLB4jmcxKCBUp1p54FtxF1wDY9H7tfwZDmDXYqClFwAvM/IqbDlNT Ch1XikbyvkYVtoO1HccalqXHxiugsZPInJIjUh+VSrfQq3a7n5HqCLXegjk3iMaEVoVo CSrXutzBjobd8eUzo8/eKoeHGfA2xxNuFMZeWE/CLrf/+AFc1kuCfKC1gli+h4HDoD93 YrjnutoHegTyZoR3nZS7+TyonKy2/PA138XxdHgnjn3i00iQttild2iUE9+dH0CDxMi+ 6ztQ== X-Gm-Message-State: AHQUAuaT7DeHOuzjOA1odkMY+uZXUcTrnLQ0Eh0Qf/6vSlLWg/RNYQMj XCOSR2P+zGqNw584chn6Jg4LwjohTjRrxh1bs6o= X-Google-Smtp-Source: AHgI3IYytXCDLwVNLvWaXUt5hiE+tLvNrkQSyU+uJN7H717iYjaYn3AW+aztWkBHflpSqJH4HE86liDlnm2z5SKqsjE= X-Received: by 2002:a24:570a:: with SMTP id u10mr8310130ita.11.1549271902531; Mon, 04 Feb 2019 01:18:22 -0800 (PST) MIME-Version: 1.0 References: <9E084747-CD0E-4729-8EFF-8625D771ED6B@gmail.com> In-Reply-To: <9E084747-CD0E-4729-8EFF-8625D771ED6B@gmail.com> From: =?UTF-8?Q?Jonas_M=C3=A5rtensson?= Date: Mon, 4 Feb 2019 10:18:11 +0100 Message-ID: To: Jonathan Morton Cc: Cake List Content-Type: multipart/alternative; boundary="00000000000069339605810df914" Subject: Re: [Cake] separate shaping of wifi 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: Mon, 04 Feb 2019 09:18:23 -0000 --00000000000069339605810df914 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable > Aside from fixing the end device so it isn't bloated in the first place, I think you have a reasonable solution there. You might only need to limit the ingress rate on wlan0, if your router's wifi stack is debloated. Right, only limit the ingress rate, that's what I'm doing and it seems to be working but I haven't seen this type of setup being discussed anywhere. The tricky part is choosing the shaping rate since the actual wifi rate varies. There's not so much I can do to fix bloat on windows and android devices. /Jonas On Sun, Feb 3, 2019 at 11:08 PM Jonathan Morton wrote: > > On 4 Feb, 2019, at 12:04 am, Jonas M=C3=A5rtensson < > martensson.jonas@gmail.com> wrote: > > > > I'm running OpenWrt with sqm on my home router. Is there any potential > problem with enabling sqm with cake on both eth0 (wan) and wlan0? The > reason for doing this is that if I only do shaping on the wan interface I > still get bad uplink bufferbloat on wifi. I assume this is because the > bottleneck in this case is actually the wifi and the bloat is in the end > device (laptop, phone, etc.). By shaping to a rate lower than the wifi > rate, the bloat disappears. Is there any better way to accomplish this th= an > to enable a second sqm instance on wlan0 if I don't want to sacrifice spe= ed > for wired devices? > > Aside from fixing the end device so it isn't bloated in the first place, = I > think you have a reasonable solution there. You might only need to limit > the ingress rate on wlan0, if your router's wifi stack is debloated. > > - Jonathan Morton > > --00000000000069339605810df914 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
> Aside from fixing the end device so = it isn't bloated in the first place, I think you have a reasonable solu= tion there.=C2=A0 You might only need to limit the ingress rate on wlan0, i= f your router's wifi stack is debloated.

Right, only limit the ingress rate, that's what I'm doing an= d it seems to be working but I haven't seen this type of setup being di= scussed anywhere. The tricky part is choosing the shaping rate since the ac= tual wifi rate varies.

There's not so much I c= an do to fix bloat on windows and android devices.

/Jonas

On Sun, Feb 3, 2019 at 11:08 PM Jonathan Morton <chromatix99@gmail.com> wrote:
> On 4 Feb, 2019, at 12:04= am, Jonas M=C3=A5rtensson <martensson.jonas@gmail.com> wrote:
>
> I'm running OpenWrt with sqm on my home router. Is there any poten= tial problem with enabling sqm with cake on both eth0 (wan) and wlan0? The = reason for doing this is that if I only do shaping on the wan interface I s= till get bad uplink bufferbloat on wifi. I assume this is because the bottl= eneck in this case is actually the wifi and the bloat is in the end device = (laptop, phone, etc.). By shaping to a rate lower than the wifi rate, the b= loat disappears. Is there any better way to accomplish this than to enable = a second sqm instance on wlan0 if I don't want to sacrifice speed for w= ired devices?

Aside from fixing the end device so it isn't bloated in the first place= , I think you have a reasonable solution there.=C2=A0 You might only need t= o limit the ingress rate on wlan0, if your router's wifi stack is deblo= ated.

=C2=A0- Jonathan Morton

--00000000000069339605810df914--