From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 16:39:43 2005 Return-Path: X-Original-To: freebsd-performance@freebsd.org Delivered-To: freebsd-performance@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7153E16A41F for ; Thu, 10 Nov 2005 16:39:43 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.193]) by mx1.FreeBSD.org (Postfix) with ESMTP id D186743D45 for ; Thu, 10 Nov 2005 16:39:42 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so297785wxc for ; Thu, 10 Nov 2005 08:39:42 -0800 (PST) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=NuOx7TaHJp9QMAs6LoPMYW1iSHGzG09Fvoqt75mokeC/r0EGMqCqFMD3jMyFY3AiCRHGtn22/SSatFaHyEZTBUQMzDlqxGJggBFxlcwOycAnpWpln/AWfMUasCr1JJfKaxFw10rTjqrGZohzzEFGTzIRv7JNdSVKXSCxTBy9muQ= Received: by 10.70.132.9 with SMTP id f9mr813323wxd; Thu, 10 Nov 2005 05:57:32 -0800 (PST) Received: by 10.70.9.10 with HTTP; Thu, 10 Nov 2005 05:57:31 -0800 (PST) Message-ID: <70e8236f0511100557y1b470a3aiaf84d2d883676bd8@mail.gmail.com> Date: Thu, 10 Nov 2005 13:57:31 +0000 From: Joao Barros To: Robert Watson In-Reply-To: <70e8236f0511100231k67d8c940p2a1875d0c09f406d@mail.gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> <20051109060931.GD5188@obiwan.tataz.chchile.org> <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> <70e8236f0511090627p24c90400ke39bdb0da222a323@mail.gmail.com> <4372184C.10809@mebius.lv> <43727712.4020500@rogers.com> <20051110090429.J33260@fledge.watson.org> <70e8236f0511100231k67d8c940p2a1875d0c09f406d@mail.gmail.com> Cc: Mike Jakubik , freebsd-performance@freebsd.org, Arkadi Shishlov Subject: Re: Poor Samba throughput on 6.0 X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 10 Nov 2005 16:39:43 -0000 On 11/10/05, Joao Barros wrote: > On 11/10/05, Robert Watson wrote: > > On Wed, 9 Nov 2005, Mike Jakubik wrote: > > > > > I have done many tests to try to determine the poor performance on my > > > systems (FreeBSD-current connected directly to Windows XP via identic= al > > > Intel Pro 1000 cards) and my only conclusion is that Samba on FreeBSD > > > when talking to a Windows box is simply fubared. Im willing to bet th= ats > > > not the case on Linux. I can get great speeds (~30MB/s) using ftp > > > between the machines, but when i use Samba, i get a maximum of 11MB/s= . I > > > have tried playing with mtu size, samba options, polling, but there d= oes > > > not seem to be any visible bottlenecks, its just simply slow. It woul= d > > > be great if someone could get to the bottom of this, its quite > > > frustrating. > > > > There are a number of TCP related configuration frobs on FreeBSD. It > > would be quite interesting to know how modifying each of the following > > settings affects Samba performance: > > > > net.inet.tcp.delayed_ack > > net.inet.tcp.sack.enable > > net.inet.tcp.inflight.enable > > > > There has been recent evidence that the inflight bandwidth calculation = is > > having problems generating stable performance at high bandwidth and low > > latency, so I might try that one first. > > > > Robert N M Watson > > I just tested on my home machine and delayed_ack and sack.enable have > little or no effect on the results. > inflight.enable on the other hand had a huge impact! With the file > cached I've gone from flat 5.5MB/s up to also flat 7.2MB/s > It's an improvement but there is still a difference up to those > theoretical 12.5MB/s on a 100mbit link > -- > Joao Barros I tested on my pet server at work and gone from 58% to 80% bandwith usage during a cached read with net.inet.tcp.inflight.enable disabled. -- Joao Barros