From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-ob0-x232.google.com (mail-ob0-x232.google.com [IPv6:2607:f8b0:4003:c01::232]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority G2" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 519FC21F563; Wed, 17 Jun 2015 17:03:12 -0700 (PDT) Received: by obbkn5 with SMTP id kn5so15173703obb.0; Wed, 17 Jun 2015 17:03:12 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:date:message-id:subject:from:to:content-type :content-transfer-encoding; bh=CGYOE9AJ7njiCvoSO3juV/crEmZ6HFPsqU+7+RdMsAM=; b=fhkcU8DldEi2w0UzeQA9qZvYUkUILKRd4nslbIXbEdlVLDflOmV2uDpdbvPmF9TyeI TPPEkj9amA/M0kZjed09eKMTOZ8c4dDaUSxcKeW0YjRqQXaGUvxi9ZjzYN0i9NyrlKR1 7JYIBK28ivtUes6qXrgekZ/4tGJo/TMyRxS9ZqntADArJwZj1hudDgi/lNsfF0GuTdIk 5mSrRb9h/vmMPkS/hkb2l+8y7V3qmiPSHnMn/50Ig/U7hfDI4uIDQYs1aXPC1pBpOfy1 NjJKRJgKw3iopK7xDPObfgCWZVWw4DsHIKJ9GdIfSDAhkEjcBkzlzlZeo2vBFQOQ9wXz cXcg== MIME-Version: 1.0 X-Received: by 10.182.88.201 with SMTP id bi9mr6814100obb.29.1434585792071; Wed, 17 Jun 2015 17:03:12 -0700 (PDT) Received: by 10.202.105.129 with HTTP; Wed, 17 Jun 2015 17:03:12 -0700 (PDT) Date: Wed, 17 Jun 2015 17:03:12 -0700 Message-ID: From: Dave Taht To: cake@lists.bufferbloat.net, "cerowrt-devel@lists.bufferbloat.net" Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: [Cerowrt-devel] sysctl for default qdisc... flakiness X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 18 Jun 2015 00:03:41 -0000 one of the issues the fedora folk ran into was that setting something like: net.core.default_qdisc=3Dcake does not always work (unless some set of circumstances are met at boot, which included compiling in the right qdisc rather than as a module) But I had thought that once that sysctl was set, if you added and removed a qdisc, the right thing would happen on mq devices - which didn't work (openwrt has an out of tree patch to change the default qdisc to fq_codel) I see this working properly on an x86 box. I guess I gotta recreate debloat.sh on this at some point, although the prospect of replacing "mq" entirely on wifi with something like cake has a great deal of appeal. --=20 Dave T=C3=A4ht worldwide bufferbloat report: http://www.dslreports.com/speedtest/results/bufferbloat And: What will it take to vastly improve wifi for everyone? https://plus.google.com/u/0/explore/makewififast