From owner-freebsd-arm@freebsd.org Tue Oct 6 22:14:11 2015 Return-Path: Delivered-To: freebsd-arm@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 9D9CF9B6179 for ; Tue, 6 Oct 2015 22:14:11 +0000 (UTC) (envelope-from garmitage@swin.edu.au) Received: from gpo3.cc.swin.edu.au (gpo3.cc.swin.edu.au [136.186.1.32]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 1C67AB1E for ; Tue, 6 Oct 2015 22:14:10 +0000 (UTC) (envelope-from garmitage@swin.edu.au) Received: from [136.186.229.37] (garmitage.caia.swin.edu.au [136.186.229.37]) by gpo3.cc.swin.edu.au (8.14.3/8.14.3) with ESMTP id t96ME6hG005181 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES128-SHA bits=128 verify=NO); Wed, 7 Oct 2015 09:14:07 +1100 Subject: Re: dwc on banana pi pro and poor network performance To: freebsd-arm@freebsd.org References: <560ED8DF.4080709@gmx.de> <560F2706.9@toomeek.waw.pl> From: grenville armitage Message-ID: <561447AE.3090509@swin.edu.au> Date: Wed, 7 Oct 2015 09:14:06 +1100 User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:38.0) Gecko/20100101 Thunderbird/38.1.0 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: freebsd-arm@freebsd.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: "Porting FreeBSD to ARM processors." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Oct 2015 22:14:11 -0000 On 10/03/2015 15:41, jau789@gmail.com wrote: > Since iperf uses tcp by default, the congestion control methods > used on the test systems may affect the test results quite seriously. > Most of the cc methods known to freebsd were not designed with > short high capasity links in mind. Many of those cc methods were > in fact intended for long distances and high bandwidth. In this case > the RTT is very short while many of the CC methods expext much > higher RTTs. I'm curious which of FreeBSD's cc_* algorithms you consider to _expect_ "much higher RTTs". This seems like something we should fix. cheers, gja > For the best results on short distances and high > bandwidth enable ECN bits and use dctcp. The dctcp method was > from day one intended to behave gracefully also when RTT is very > short and the bandwidth is large. > > --jau [...]