From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from complete.lackof.org (complete.lackof.org [198.49.126.79]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "mail.lackof.org", Issuer "CAcert Class 3 Root" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id DB59921F1A1 for ; Fri, 17 Oct 2014 14:06:43 -0700 (PDT) Received: from taggart.lackof.org (c-98-203-137-247.hsd1.wa.comcast.net [98.203.137.247]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "taggart.lackof.org", Issuer "CAcert Class 3 Root" (verified OK)) by complete.lackof.org (Postfix) with ESMTPS id D165C33E01AD for ; Fri, 17 Oct 2014 15:06:41 -0600 (MDT) Received: by taggart.lackof.org (Postfix, from userid 1000) id 1615F1CD; Fri, 17 Oct 2014 14:06:41 -0700 (PDT) Received: from zorak.home.bogus (localhost [127.0.0.1]) by taggart.lackof.org (Postfix) with ESMTP id 15163148 for ; Fri, 17 Oct 2014 14:06:41 -0700 (PDT) X-Mailer: exmh version 2.8.0 04/21/2012 (debian 1:2.8.0~rc1-2) with nmh-1.5 From: Matt Taggart To: cerowrt-devel@lists.bufferbloat.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 17 Oct 2014 14:06:41 -0700 Message-Id: <20141017210641.1615F1CD@taggart.lackof.org> X-Virus-Scanned: clamav-milter 0.98.1 at complete.lackof.org X-Virus-Status: Clean X-Spam-Status: No, score=0.0 required=5.0 tests=UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on complete.lackof.org Subject: [Cerowrt-devel] fq_codel tuning on distros 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: Fri, 17 Oct 2014 21:07:12 -0000 Hi, http://www.bufferbloat.net/projects/codel/wiki/Best_practices_for_benchmarki ng_Codel_and_FQ_Codel#Tuning-fq_codel explains that the default packet limit of 10000 is designed for 10GigE speeds and that for slower links it should be turned down. Is that still true? Looking at 3.16 source in fq_codel_init I see: sch->limit = 10*1024; q->flows_cnt = 1024; But I don't know what those correspond to. Are they sysfs tunable or only at compile time? If Linux distros are going to turn on fq_codel by default, are these reasonable values for the installed base (which I am assuming is mostly 1GigE)? What recommendations should the distro documentation make for tuning on various speeds? I'm excited for this to go into distros, what needs to be done to make that easier? Thanks, -- Matt Taggart matt@lackof.org