From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 02:48:38 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 2B84D16A41F for ; Wed, 9 Nov 2005 02:48:38 +0000 (GMT) (envelope-from mv@roq.com) Received: from p4.roq.com (ns1.ecoms.com [207.44.130.137]) by mx1.FreeBSD.org (Postfix) with ESMTP id D17AC43D48 for ; Wed, 9 Nov 2005 02:48:37 +0000 (GMT) (envelope-from mv@roq.com) Received: from p4.roq.com (localhost.roq.com [127.0.0.1]) by p4.roq.com (Postfix) with ESMTP id C29DA4CE68; Wed, 9 Nov 2005 02:48:49 +0000 (GMT) Received: from [192.168.46.52] (ppp166-27.static.internode.on.net [150.101.166.27]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by p4.roq.com (Postfix) with ESMTP id DC33C4CE09; Wed, 9 Nov 2005 02:48:48 +0000 (GMT) Message-ID: <43716381.4010601@roq.com> Date: Wed, 09 Nov 2005 13:48:33 +1100 From: Michael Vince User-Agent: Mozilla/5.0 (X11; U; FreeBSD amd64; en-US; rv:1.7.12) Gecko/20051019 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Joao Barros References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> In-Reply-To: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: ClamAV using ClamSMTP Cc: freebsd-performance@freebsd.org 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: Wed, 09 Nov 2005 02:48:38 -0000 Joao Barros wrote: >Hi, > >Last month I started a thread[1] on current@ about this, but I guess I >should have done it here, my apologies for that. > >After my initial post I did some more testing and I'm going to start >clean here with all my findings :) > >I started with Samba 3 installed on a PIII 733MHz with fxp (82559) and >a RAID5 consisting of 4 drives connected to an amr. >Performance reading or writing was poor, around 5.5MB/s measured on >two Windows clients and iostat never topped that by much. >cpu was mbufs were available and there were no IRQs shared. >To dismiss the amr out of the question I tried with a local IDE >attached yielding the same results. >I then tested the same on a machine I have at work, an HP Proliant >server, Pentium 4 3.06GHz, used SMP instead of GENERIC to use HTT. >I could get 8MB/s with 2 read or write simultaneous operations. With 1 >operation I still can only get 6MB/s >This machine has 1GB ram and after copying a 700MB file to it it was all cached. >A copy to dev/null took 1 second. >A copy via samba took the same time as if there was no cache for it. >iostat always showed 0.0 during the operation so that pretty much >takes disks, controllers, IO out of the picture. > >Both machines have cpu, IO and mbufs to spare and they still can't use >them. Why? > > >[1] http://lists.freebsd.org/pipermail/freebsd-current/2005-October/057116.html > >-- >Joao Barros >_______________________________________________ > > Sometime in the near future I will be building a Samba3 server and I plan to get everything I can get out of it, I expect to uncover a lot of needed tweaks to get it going fast. My guess would be to you try at least turning on polling, also if its only 100mbit Ethernet/switches you got then I guess you cant expect much either. Mike From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 06:09:55 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 07C9016A41F for ; Wed, 9 Nov 2005 06:09:55 +0000 (GMT) (envelope-from tataz@tataz.chchile.org) Received: from smtp5-g19.free.fr (smtp5-g19.free.fr [212.27.42.35]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9CED343D45 for ; Wed, 9 Nov 2005 06:09:54 +0000 (GMT) (envelope-from tataz@tataz.chchile.org) Received: from tatooine.tataz.chchile.org (vol75-8-82-233-239-98.fbx.proxad.net [82.233.239.98]) by smtp5-g19.free.fr (Postfix) with ESMTP id 2815895E0; Wed, 9 Nov 2005 07:09:53 +0100 (CET) Received: by tatooine.tataz.chchile.org (Postfix, from userid 1000) id 3F5AE4083; Wed, 9 Nov 2005 07:09:31 +0100 (CET) Date: Wed, 9 Nov 2005 07:09:31 +0100 From: Jeremie Le Hen To: Joao Barros Message-ID: <20051109060931.GD5188@obiwan.tataz.chchile.org> References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> User-Agent: Mutt/1.5.11 Cc: freebsd-performance@freebsd.org 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: Wed, 09 Nov 2005 06:09:55 -0000 Hi, Joao, > Last month I started a thread[1] on current@ about this, but I guess I > should have done it here, my apologies for that. > > After my initial post I did some more testing and I'm going to start > clean here with all my findings :) > > I started with Samba 3 installed on a PIII 733MHz with fxp (82559) and > a RAID5 consisting of 4 drives connected to an amr. > Performance reading or writing was poor, around 5.5MB/s measured on > two Windows clients and iostat never topped that by much. > cpu was mbufs were available and there were no IRQs shared. > To dismiss the amr out of the question I tried with a local IDE > attached yielding the same results. > I then tested the same on a machine I have at work, an HP Proliant > server, Pentium 4 3.06GHz, used SMP instead of GENERIC to use HTT. > I could get 8MB/s with 2 read or write simultaneous operations. With 1 > operation I still can only get 6MB/s > This machine has 1GB ram and after copying a 700MB file to it it was > all cached. > A copy to dev/null took 1 second. > A copy via samba took the same time as if there was no cache for it. > iostat always showed 0.0 during the operation so that pretty much > takes disks, controllers, IO out of the picture. > > Both machines have cpu, IO and mbufs to spare and they still can't use > them. Why? I won't be able to help you much, but as almost nobody answered you, I take it for the moment in order to ask you some more informations. Which scheduler are you using, 4BSD or ULE ? It might be worth testing the other one and sending us the new benchmark results. Also, if you are able to remove a drive from your RAID5, you can try R/W performances from/to it, without using amr(4), both with 4BSD and ULE. Regards, -- Jeremie Le Hen < jeremie at le-hen dot org >< ttz at chchile dot org > From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 14:54:54 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 EB64C16A420 for ; Wed, 9 Nov 2005 14:54:53 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.194]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0FFB743D67 for ; Wed, 9 Nov 2005 14:54:49 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so279990wxc for ; Wed, 09 Nov 2005 06:54:49 -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=Hf5x4qWDCdbREjIO8V8XKUlWI/D0iU00dheJKVyasZ0EiWPIjbr6yNJ8PLk+3t8S0TO36gPxURIXENwTIW+JN4hX5eO1OBrT7qdYZfvZj1nJEcETJJNnlz4Mr0idfKr52cNZNvVBzM+vXoIpdmHWQDNhNgEiPLC3T1jKqofGgG0= Received: by 10.70.47.20 with SMTP id u20mr848431wxu; Wed, 09 Nov 2005 06:27:32 -0800 (PST) Received: by 10.70.9.10 with HTTP; Wed, 9 Nov 2005 06:27:32 -0800 (PST) Message-ID: <70e8236f0511090627p24c90400ke39bdb0da222a323@mail.gmail.com> Date: Wed, 9 Nov 2005 14:27:32 +0000 From: Joao Barros To: Jeremie Le Hen In-Reply-To: <70e8236f0511090517g29b3f887x1b97ef5dec04548@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> Cc: freebsd-performance@freebsd.org 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: Wed, 09 Nov 2005 14:54:54 -0000 On 11/9/05, Joao Barros wrote: > On 11/9/05, Jeremie Le Hen wrote: > > Hi, Joao, > > > > > Last month I started a thread[1] on current@ about this, but I guess = I > > > should have done it here, my apologies for that. > > > > > > After my initial post I did some more testing and I'm going to start > > > clean here with all my findings :) > > > > > > I started with Samba 3 installed on a PIII 733MHz with fxp (82559) an= d > > > a RAID5 consisting of 4 drives connected to an amr. > > > Performance reading or writing was poor, around 5.5MB/s measured on > > > two Windows clients and iostat never topped that by much. > > > cpu was mbufs were available and there were no IRQs shared. > > > To dismiss the amr out of the question I tried with a local IDE > > > attached yielding the same results. > > > I then tested the same on a machine I have at work, an HP Proliant > > > server, Pentium 4 3.06GHz, used SMP instead of GENERIC to use HTT. > > > I could get 8MB/s with 2 read or write simultaneous operations. With = 1 > > > operation I still can only get 6MB/s > > > This machine has 1GB ram and after copying a 700MB file to it it was > > > all cached. > > > A copy to dev/null took 1 second. > > > A copy via samba took the same time as if there was no cache for it. > > > iostat always showed 0.0 during the operation so that pretty much > > > takes disks, controllers, IO out of the picture. > > > > > > Both machines have cpu, IO and mbufs to spare and they still can't us= e > > > them. Why? > > > > I won't be able to help you much, but as almost nobody answered you, > > I take it for the moment in order to ask you some more informations. > > > > Which scheduler are you using, 4BSD or ULE ? It might be worth testing > > the other one and sending us the new benchmark results. > > The testings were all with either GENERIC or SMP thus using 4BSD, I > can try ULE and see if I get any different results. > On a P4 3.06GHz with HTT enabled and ULE I get the same results. I get a flat line at 58% looking at the bandwith in task manager on a Windows 2003 Server while doing a cached read. I can get up to 70% bandwith during writes. Percentages are relative to 100Mbits bandwith. > > > > Also, if you are able to remove a drive from your RAID5, you can try > > R/W performances from/to it, without using amr(4), both with 4BSD and > > ULE. > > I tried using a single drive, an IDE and a SCSI-2 and on 2 machines at > work both with a RAID1. > Even better, there is a part in my initial email where I mention that > having a 700MB file cached (iostat reported no reads) the results were > the same. With this in mind I don't think the problem is at the > storage level. -- Joao Barros From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 15:40:53 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 A324F16A41F for ; Wed, 9 Nov 2005 15:40:53 +0000 (GMT) (envelope-from arkadi@mebius.lv) Received: from mail.hosting.lv (mail.hosting.lv [62.85.37.83]) by mx1.FreeBSD.org (Postfix) with ESMTP id EC34943D66 for ; Wed, 9 Nov 2005 15:40:45 +0000 (GMT) (envelope-from arkadi@mebius.lv) Received: from lithium.bad.lv ([62.85.6.146]) by mail.hosting.lv with esmtps (TLSv1:AES256-SHA:256) id 1EZs3P-000G6A-7Q; Wed, 09 Nov 2005 17:40:08 +0200 Message-ID: <4372184C.10809@mebius.lv> Date: Wed, 09 Nov 2005 17:39:56 +0200 From: Arkadi Shishlov Organization: Mebius IT User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.12) Gecko/20051001 X-Accept-Language: en MIME-Version: 1.0 To: freebsd-performance@freebsd.org References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> <20051109060931.GD5188@obiwan.tataz.chchile.org> <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> <70e8236f0511090627p24c90400ke39bdb0da222a323@mail.gmail.com> In-Reply-To: <70e8236f0511090627p24c90400ke39bdb0da222a323@mail.gmail.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Spam-Score: -2.8 (--) 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: Wed, 09 Nov 2005 15:40:53 -0000 Joao Barros wrote: > On a P4 3.06GHz with HTT enabled and ULE I get the same results. > I get a flat line at 58% looking at the bandwith in task manager on a > Windows 2003 Server while doing a cached read. > I can get up to 70% bandwith during writes. > Percentages are relative to 100Mbits bandwith. Are you able to get better throughput from different version of FreeBSD or different OS (Linux) with Samba? From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 22:24:14 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 8FBAE16A41F for ; Wed, 9 Nov 2005 22:24:14 +0000 (GMT) (envelope-from mikej@rogers.com) Received: from smtp101.rog.mail.re2.yahoo.com (smtp101.rog.mail.re2.yahoo.com [206.190.36.79]) by mx1.FreeBSD.org (Postfix) with SMTP id 0D85F43D48 for ; Wed, 9 Nov 2005 22:24:11 +0000 (GMT) (envelope-from mikej@rogers.com) Received: (qmail 20556 invoked from network); 9 Nov 2005 22:24:10 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=rogers.com; h=Received:Message-ID:Date:From:User-Agent:MIME-Version:To:CC:Subject:References:In-Reply-To:Content-Type:Content-Transfer-Encoding; b=IHMFt2U14/Sm+DtMrY9DwV9aZgQW9qz+bSjfTlYzeQQoana2P9PbuUgRATTTCo3fkDbbD4kCXzHR74d5BbcE6kklhsPR74WHFIyeusslhrhdN+IsYPdG35xhKpqxtaCXnTuZ0V2nhzXg0vFOH5L5rgA8hmajjKA/ok46RWxuf58= ; Received: from unknown (HELO ?172.16.0.199?) (mikej@rogers.com@70.31.50.81 with plain) by smtp101.rog.mail.re2.yahoo.com with SMTP; 9 Nov 2005 22:24:10 -0000 Message-ID: <43727712.4020500@rogers.com> Date: Wed, 09 Nov 2005 17:24:18 -0500 From: Mike Jakubik User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: Arkadi Shishlov References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> <20051109060931.GD5188@obiwan.tataz.chchile.org> <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> <70e8236f0511090627p24c90400ke39bdb0da222a323@mail.gmail.com> <4372184C.10809@mebius.lv> In-Reply-To: <4372184C.10809@mebius.lv> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-performance@freebsd.org 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: Wed, 09 Nov 2005 22:24:14 -0000 Arkadi Shishlov wrote: > Joao Barros wrote: > >> On a P4 3.06GHz with HTT enabled and ULE I get the same results. >> I get a flat line at 58% looking at the bandwith in task manager on a >> Windows 2003 Server while doing a cached read. >> I can get up to 70% bandwith during writes. >> Percentages are relative to 100Mbits bandwith. >> > > Are you able to get better throughput from different version of FreeBSD or > different OS (Linux) with Samba? > I have done many tests to try to determine the poor performance on my systems (FreeBSD-current connected directly to Windows XP via identical 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 thats 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 does not seem to be any visible bottlenecks, its just simply slow. It would be great if someone could get to the bottom of this, its quite frustrating. From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 22:36:38 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 11C2F16A41F for ; Wed, 9 Nov 2005 22:36:38 +0000 (GMT) (envelope-from nevans@talkpoint.com) Received: from relay.talkpoint.com (pobox.talkpoint.com [204.141.15.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9A3A543D45 for ; Wed, 9 Nov 2005 22:36:37 +0000 (GMT) (envelope-from nevans@talkpoint.com) Received: from ASSP-nospam ([127.0.0.1]) by relay.talkpoint.com with Microsoft SMTPSVC(5.0.2195.6713); Wed, 9 Nov 2005 17:36:36 -0500 Received: from 204.141.15.194 ([204.141.15.194] helo=pleiades.nextvenue.com) by ASSP-nospam ; 9 Nov 05 22:36:36 -0000 Date: Wed, 9 Nov 2005 17:36:34 -0500 From: Nick Evans To: Mike Jakubik Message-ID: <20051109173634.22391fec@pleiades.nextvenue.com> In-Reply-To: <43727712.4020500@rogers.com> 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> X-Mailer: Sylpheed-Claws 1.9.15 (GTK+ 2.6.10; i386-portbld-freebsd6.0) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-OriginalArrivalTime: 09 Nov 2005 22:36:36.0603 (UTC) FILETIME=[0B2C74B0:01C5E57E] Cc: 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: Wed, 09 Nov 2005 22:36:38 -0000 On Wed, 09 Nov 2005 17:24:18 -0500 Mike Jakubik wrote: > Arkadi Shishlov wrote: > > Joao Barros wrote: > > > >> On a P4 3.06GHz with HTT enabled and ULE I get the same results. > >> I get a flat line at 58% looking at the bandwith in task manager on a > >> Windows 2003 Server while doing a cached read. > >> I can get up to 70% bandwith during writes. > >> Percentages are relative to 100Mbits bandwith. > >> > > > > Are you able to get better throughput from different version of FreeBSD or > > different OS (Linux) with Samba? > > > > I have done many tests to try to determine the poor performance on my > systems (FreeBSD-current connected directly to Windows XP via identical > 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 thats > 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 does > not seem to be any visible bottlenecks, its just simply slow. It would > be great if someone could get to the bottom of this, its quite frustrating. > Has anyone tried mounting a FreeBSD/Samba share with smbfs from another FreeBSD machine? Also are the Windows PC's stock or have you tried tweaking TCP there? From owner-freebsd-performance@FreeBSD.ORG Wed Nov 9 23:23:46 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 7DD5B16A41F for ; Wed, 9 Nov 2005 23:23:46 +0000 (GMT) (envelope-from killing@multiplay.co.uk) Received: from multiplay.co.uk (core6.multiplay.co.uk [85.236.96.23]) by mx1.FreeBSD.org (Postfix) with ESMTP id B3AB043D46 for ; Wed, 9 Nov 2005 23:23:45 +0000 (GMT) (envelope-from killing@multiplay.co.uk) Received: from vader ([212.135.219.179]) by multiplay.co.uk (multiplay.co.uk [85.236.96.23]) (MDaemon.PRO.v8.1.3.R) with ESMTP id md50001998048.msg for ; Wed, 09 Nov 2005 23:22:48 +0000 Message-ID: <002601c5e584$7f638700$b3db87d4@multiplay.co.uk> From: "Steven Hartland" To: "Nick Evans" , "Mike Jakubik" 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> <20051109173634.22391fec@pleiades.nextvenue.com> Date: Wed, 9 Nov 2005 23:22:42 -0000 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2670 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2670 X-Spam-Processed: multiplay.co.uk, Wed, 09 Nov 2005 23:22:48 +0000 (not processed: message from valid local sender) X-MDRemoteIP: 212.135.219.179 X-Return-Path: killing@multiplay.co.uk X-MDaemon-Deliver-To: freebsd-performance@freebsd.org X-MDAV-Processed: multiplay.co.uk, Wed, 09 Nov 2005 23:22:50 +0000 Cc: 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: Wed, 09 Nov 2005 23:23:46 -0000 Just did a few quick tests on 5.4 here ( not upgraded to 6.0 yet ) and on Gig I get a max of 20Mb/s using samba with the following options: socket options = TCP_NODELAY SO_RCVBUF=131072 SO_SNDBUF=131072 max xmit = 131072 With ftp I can get 45Mb/s Steve ================================================ This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137 or return the E.mail to postmaster@multiplay.co.uk. From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 02:00:45 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 2D84316A41F for ; Thu, 10 Nov 2005 02:00:45 +0000 (GMT) (envelope-from mikej@rogers.com) Received: from smtp102.rog.mail.re2.yahoo.com (smtp102.rog.mail.re2.yahoo.com [206.190.36.80]) by mx1.FreeBSD.org (Postfix) with SMTP id 9A0B543D48 for ; Thu, 10 Nov 2005 02:00:44 +0000 (GMT) (envelope-from mikej@rogers.com) Received: (qmail 59817 invoked from network); 10 Nov 2005 02:00:40 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=rogers.com; h=Received:Message-ID:Date:From:User-Agent:MIME-Version:To:CC:Subject:References:In-Reply-To:Content-Type:Content-Transfer-Encoding; b=bOklKvTTu/McE/dGbEssCgota5aEjUQU3TihhPryFpzU0Ju/QYYmQsZ/6M0uBmztKTdzhkTDjjjeWOV5ykAk4M2myh75lF92Od3gWbKL1MRUeX0Jjgyi2658TlxnGnDr0kSM6OIk+Efd9ZyoFSbTcqCZY3wqhBlkIjrkwOl6aWY= ; Received: from unknown (HELO ?172.16.0.199?) (mikej@rogers.com@70.31.50.81 with plain) by smtp102.rog.mail.re2.yahoo.com with SMTP; 10 Nov 2005 02:00:40 -0000 Message-ID: <4372A9C9.2090803@rogers.com> Date: Wed, 09 Nov 2005 21:00:41 -0500 From: Mike Jakubik User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: Nick Evans 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> <20051109173634.22391fec@pleiades.nextvenue.com> In-Reply-To: <20051109173634.22391fec@pleiades.nextvenue.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: 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 02:00:45 -0000 Nick Evans wrote: > On Wed, 09 Nov 2005 17:24:18 -0500 > 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 identical >> 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 thats >> 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 does >> not seem to be any visible bottlenecks, its just simply slow. It would >> be great if someone could get to the bottom of this, its quite frustrating. >> >> > > Has anyone tried mounting a FreeBSD/Samba share with smbfs from another > FreeBSD machine? Also are the Windows PC's stock or have you tried tweaking > TCP there? > Yes, here are my results, WindowsXP is stock. Machine A ----------- FreeBSD 7.0-CURRENT CPU: AMD Athlon(tm) Processor (1410.21-MHz 686-class CPU) Mem: 512MB em0: Samba-3.0.20b,1 server Machine B ----------- Dual boot FreeBSD 6R and WndowsXP CPU: AthlonXP 2400 Mem: 1GB em0: Samba client Tests sending a random 400MB file: A sending to B(FreeBSD) -------------------------- Samba = 13MB/s Ftp = 16MB/s A sending to B(WindowsXP) ----------------------------- Samba = 13MB Ftp = 16MB B(FreeBSD) sending to A -------------------------- Samba = 2.5MB/s Ftp = 25MB/s B(WindowsXP) sending to A ----------------------------- Samba = 17MB/s Ftp = 31MB/s Very, very odd results. Machine B has raid0, yet it experiences the slowest incoming speeds in ftp. I cant begin to imagine where the problem is, but we can see that there is no real hardware bottlenecks. From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 09:06:37 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 491C216A41F for ; Thu, 10 Nov 2005 09:06:37 +0000 (GMT) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [209.31.154.42]) by mx1.FreeBSD.org (Postfix) with ESMTP id EEAE643D45 for ; Thu, 10 Nov 2005 09:06:36 +0000 (GMT) (envelope-from rwatson@FreeBSD.org) Received: from fledge.watson.org (fledge.watson.org [209.31.154.41]) by cyrus.watson.org (Postfix) with ESMTP id 970B546B1C; Thu, 10 Nov 2005 04:06:35 -0500 (EST) Date: Thu, 10 Nov 2005 09:06:35 +0000 (GMT) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Mike Jakubik In-Reply-To: <43727712.4020500@rogers.com> Message-ID: <20051110090429.J33260@fledge.watson.org> 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> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: 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 09:06:37 -0000 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 identical > 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 thats > 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 does > not seem to be any visible bottlenecks, its just simply slow. It would > 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 From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 12:48:14 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 BF74A16A421 for ; Thu, 10 Nov 2005 12:48:14 +0000 (GMT) (envelope-from mikej@rogers.com) Received: from smtp103.rog.mail.re2.yahoo.com (smtp103.rog.mail.re2.yahoo.com [206.190.36.81]) by mx1.FreeBSD.org (Postfix) with SMTP id C74DB43D53 for ; Thu, 10 Nov 2005 12:48:13 +0000 (GMT) (envelope-from mikej@rogers.com) Received: (qmail 14489 invoked from network); 10 Nov 2005 12:48:13 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=rogers.com; h=Received:Message-ID:Date:From:User-Agent:MIME-Version:To:CC:Subject:References:In-Reply-To:Content-Type:Content-Transfer-Encoding; b=2+vRHmP9yGsrPJ8QEHBbCcQjVKdHKVop8JsbgMrfz60VH3ucSgmvjYML8KdyCyfhQtFR+DV27TBJBi2FRt4Tz076YS+U8cj7nB/hHUc3iQvCg2J9pbmSuRACV8PLlW4NQ42jEOqG7I2CJ8O/OClhpH508Wc/83Hj8quU2HQPXpI= ; Received: from unknown (HELO ?172.16.0.199?) (mikej@rogers.com@70.31.50.81 with plain) by smtp103.rog.mail.re2.yahoo.com with SMTP; 10 Nov 2005 12:48:13 -0000 Message-ID: <4373418E.1080907@rogers.com> Date: Thu, 10 Nov 2005 07:48:14 -0500 From: Mike Jakubik User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: Robert Watson 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> In-Reply-To: <20051110090429.J33260@fledge.watson.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: 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 12:48:14 -0000 Robert Watson wrote: > 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. I get about ~3MB/s extra, and a 16MB/s peak, with inflight disabled, when downloading from the FreeBSD server to WindowsXP. An improvement, but still not a desirable result. The other variables do not seem to make much of a difference. From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 15:32:07 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 48B7F16A420; Thu, 10 Nov 2005 15:32:07 +0000 (GMT) (envelope-from nevans@talkpoint.com) Received: from relay.talkpoint.com (pobox.talkpoint.com [204.141.15.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id CBABF43D48; Thu, 10 Nov 2005 15:32:06 +0000 (GMT) (envelope-from nevans@talkpoint.com) Received: from ASSP-nospam ([127.0.0.1]) by relay.talkpoint.com with Microsoft SMTPSVC(5.0.2195.6713); Thu, 10 Nov 2005 10:32:05 -0500 Received: from 204.141.15.194 ([204.141.15.194] helo=pleiades.nextvenue.com) by ASSP-nospam ; 10 Nov 05 15:32:05 -0000 Date: Thu, 10 Nov 2005 10:32:05 -0500 From: Nick Evans To: Mike Jakubik Message-ID: <20051110103205.78d63c87@pleiades.nextvenue.com> In-Reply-To: <4373418E.1080907@rogers.com> 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> <4373418E.1080907@rogers.com> X-Mailer: Sylpheed-Claws 1.9.15 (GTK+ 2.6.10; i386-portbld-freebsd6.0) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-OriginalArrivalTime: 10 Nov 2005 15:32:05.0732 (UTC) FILETIME=[E7C3BA40:01C5E60B] Cc: freebsd-performance@freebsd.org, Robert Watson , 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 15:32:07 -0000 On Thu, 10 Nov 2005 07:48:14 -0500 Mike Jakubik wrote: > Robert Watson wrote: > > 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. > I get about ~3MB/s extra, and a 16MB/s peak, with inflight disabled, > when downloading from the FreeBSD server to WindowsXP. An improvement, > but still not a desirable result. The other variables do not seem to > make much of a difference. > _______________________________________________ > freebsd-performance@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-performance > To unsubscribe, send any mail to > "freebsd-performance-unsubscribe@freebsd.org" What's the round trip time between the two devices? You might want to look at adjusting socket buffer sizes on Windows and FreeBSD per your bandwidth delay product. Nick From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 15:59:49 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 4AB1216A41F for ; Thu, 10 Nov 2005 15:59:49 +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 4FAB843D75 for ; Thu, 10 Nov 2005 15:59:40 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so284105wxc for ; Thu, 10 Nov 2005 07:59:40 -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=WRc3P4V0C5UWVjLKAcUJqIcigVor0F+y2ZyhmR7wTPfwVKnWGF8pHTeMTYQ1nJOKmv+BqLTUAwX6EDuTNsrelHTGpY6gJxfi/bED7IJJgahInt08nEKz6Ss3vyIVnqZPUXct6nkKOHgjzeSyz31H8lfRqhuElCUBHDz2Hz7fcGQ= Received: by 10.70.33.14 with SMTP id g14mr750335wxg; Wed, 09 Nov 2005 05:13:10 -0800 (PST) Received: by 10.70.9.10 with HTTP; Wed, 9 Nov 2005 05:13:10 -0800 (PST) Message-ID: <70e8236f0511090513q2db9168by22b9ab741f8962c2@mail.gmail.com> Date: Wed, 9 Nov 2005 13:13:10 +0000 From: Joao Barros To: Michael Vince In-Reply-To: <43716381.4010601@roq.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> <43716381.4010601@roq.com> Cc: freebsd-performance@freebsd.org 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 15:59:49 -0000 On 11/9/05, Michael Vince wrote: > Joao Barros wrote: > > >Hi, > > > >Last month I started a thread[1] on current@ about this, but I guess I > >should have done it here, my apologies for that. > > > >After my initial post I did some more testing and I'm going to start > >clean here with all my findings :) > > > >I started with Samba 3 installed on a PIII 733MHz with fxp (82559) and > >a RAID5 consisting of 4 drives connected to an amr. > >Performance reading or writing was poor, around 5.5MB/s measured on > >two Windows clients and iostat never topped that by much. > >cpu was mbufs were available and there were no IRQs shared. > >To dismiss the amr out of the question I tried with a local IDE > >attached yielding the same results. > >I then tested the same on a machine I have at work, an HP Proliant > >server, Pentium 4 3.06GHz, used SMP instead of GENERIC to use HTT. > >I could get 8MB/s with 2 read or write simultaneous operations. With 1 > >operation I still can only get 6MB/s > >This machine has 1GB ram and after copying a 700MB file to it it was all= cached. > >A copy to dev/null took 1 second. > >A copy via samba took the same time as if there was no cache for it. > >iostat always showed 0.0 during the operation so that pretty much > >takes disks, controllers, IO out of the picture. > > > >Both machines have cpu, IO and mbufs to spare and they still can't use > >them. Why? > > > > > >[1] http://lists.freebsd.org/pipermail/freebsd-current/2005-October/0571= 16.html > > > >-- > >Joao Barros > >_______________________________________________ > > > > > Sometime in the near future I will be building a Samba3 server and I > plan to get everything I can get out of it, I expect to uncover a lot of > needed tweaks to get it going fast. > My guess would be to you try at least turning on polling, also if its > only 100mbit Ethernet/switches you got then I guess you cant expect much > either. > > Mike 3 years ago I had a linux machine with a less powerfull cpu maxing out the interface with SMB transfers. In all available enviroments under a 100mbit switch I can max out the interfaces using Windows or Linux. On my particular enviroment (a 3com switch at home) I can max out the interfaces doing transfers between my laptop and my home pc, both running XP. -- Joao Barros From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 16:30:41 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 9CE5316A420 for ; Thu, 10 Nov 2005 16:30:41 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.203]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2A91A43D46 for ; Thu, 10 Nov 2005 16:30:41 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so294819wxc for ; Thu, 10 Nov 2005 08:30:40 -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=L+hOiTgWZuHFjKUY+O4eruXQv1HpaOmgaaFXgINwi3QLoG6Suzu3KXUkeDyPhup+husWK3agmtqqaz1mjTqS0/30uc4GUhMGln03ERB/F9ZSem03KhkWwTNG6jcNzEK/yCFRWRbZwfKdd06GJSlnvfYpVGzhZxoFbK84/B8CVRs= Received: by 10.70.11.2 with SMTP id 2mr931446wxk; Wed, 09 Nov 2005 08:17:09 -0800 (PST) Received: by 10.70.9.10 with HTTP; Wed, 9 Nov 2005 08:17:09 -0800 (PST) Message-ID: <70e8236f0511090817v3fac7507i3cc8665bb2aa3916@mail.gmail.com> Date: Wed, 9 Nov 2005 16:17:09 +0000 From: Joao Barros To: Arkadi Shishlov In-Reply-To: <4372184C.10809@mebius.lv> 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> Cc: freebsd-performance@freebsd.org 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 16:30:41 -0000 On 11/9/05, Arkadi Shishlov wrote: > Joao Barros wrote: > > On a P4 3.06GHz with HTT enabled and ULE I get the same results. > > I get a flat line at 58% looking at the bandwith in task manager on a > > Windows 2003 Server while doing a cached read. > > I can get up to 70% bandwith during writes. > > Percentages are relative to 100Mbits bandwith. > > Are you able to get better throughput from different version of FreeBSD o= r > different OS (Linux) with Samba? I only tried FreeBSD 6 from a couple months back up to 6.0 RELEASE on this machines. The home server I first started this tests used to be a storage server via Samba running Linux and it maxed out the interfaces with little trouble. -- Joao Barros 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 From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 19:03:33 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 718EA16A41F for ; Thu, 10 Nov 2005 19:03:33 +0000 (GMT) (envelope-from brad-fbsd-perf@duttonbros.com) Received: from uno.mnl.com (uno.mnl.com [63.97.246.49]) by mx1.FreeBSD.org (Postfix) with SMTP id 662BD43D99 for ; Thu, 10 Nov 2005 19:03:06 +0000 (GMT) (envelope-from brad-fbsd-perf@duttonbros.com) Received: (qmail 62463 invoked by uid 85); 10 Nov 2005 19:03:04 -0000 Received: from 127.0.0.1 by uno (envelope-from , uid 89) with qmail-scanner-1.25 (spamassassin: 2.55. Clear:RC:1(127.0.0.1):. Processed in 0.050185 secs); 10 Nov 2005 19:03:04 -0000 Received: from unknown (HELO uno.mnl.com) (127.0.0.1) by localhost with SMTP; 10 Nov 2005 19:03:03 -0000 Received: from 67.169.82.217 (SquirrelMail authenticated user bdutton) by uno.mnl.com with HTTP; Thu, 10 Nov 2005 11:03:03 -0800 (PST) Message-ID: <54050.67.169.82.217.1131649383.squirrel@uno.mnl.com> In-Reply-To: <20051110090429.J33260@fledge.watson.org> 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> Date: Thu, 10 Nov 2005 11:03:03 -0800 (PST) From: "Bradley W. Dutton" To: freebsd-performance@freebsd.org User-Agent: SquirrelMail/1.4.5 MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Priority: 3 (Normal) Importance: Normal Subject: Re: Poor Samba throughput on 6.0 X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: brad-fbsd-perf@duttonbros.com List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 10 Nov 2005 19:03:33 -0000 I recently upgraded to gigabit and was expecting faster Samba performance as well. I don't recall the exact numbers but net.inet.tcp.delayed_ack=0 did help quite a bit on my box. I too tried mtu (intel cards and SMC jumbo switch) and polling but the transfer rates would max out around 10megs/sec with Samba. Disk transfer rate shouldn't be a problem, the disk easily does 20+meg/sec writes. ttcp transfer rates are 30+ megs/sec. One thing I did notice with ttcp is that using the -D option (TCP_NODELAY?) helped a lot. I'm using a freebsd/samba 6.0 server and windows xp clients. I'll test the other 2 sysctls over the next few days to see if they help. Thanks, Brad > 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 identical >> 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 thats >> 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 does >> not seem to be any visible bottlenecks, its just simply slow. It would >> 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 From owner-freebsd-performance@FreeBSD.ORG Thu Nov 10 19:08:27 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 64D7616A420 for ; Thu, 10 Nov 2005 19:08:27 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.201]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7FED743D77 for ; Thu, 10 Nov 2005 19:08:23 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so340466wxc for ; Thu, 10 Nov 2005 11:08:22 -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=iWl37Ni191ARO4dzgqstabcqG0ecSo+JR1VlWvr0cUQozBGJrBRqJOVU/85NVgOXE+K4tCWoNz4jx8jErAUUDXwDzMFy4QUwC/ig2nbZDga2XlXBmkzReCgNo0ltOhc3RZpG66M6cfm27ALiFgp1DQn+EyzxWW9kSuhc6f7Anok= Received: by 10.70.132.9 with SMTP id f9mr633939wxd; Thu, 10 Nov 2005 02:31:03 -0800 (PST) Received: by 10.70.9.10 with HTTP; Thu, 10 Nov 2005 02:31:03 -0800 (PST) Message-ID: <70e8236f0511100231k67d8c940p2a1875d0c09f406d@mail.gmail.com> Date: Thu, 10 Nov 2005 10:31:03 +0000 From: Joao Barros To: Robert Watson In-Reply-To: <20051110090429.J33260@fledge.watson.org> 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> 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 19:08:27 -0000 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 identical > > 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 that= s > > 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 doe= s > > not seem to be any visible bottlenecks, its just simply slow. It would > > 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 From owner-freebsd-performance@FreeBSD.ORG Fri Nov 11 04:21:12 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 3938916A41F for ; Fri, 11 Nov 2005 04:21:12 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.200]) by mx1.FreeBSD.org (Postfix) with ESMTP id BC6BB43D46 for ; Fri, 11 Nov 2005 04:21:11 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so452510wxc for ; Thu, 10 Nov 2005 20:21:11 -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=ZRHchPVBN+QBdq0eYWKgWa8gW5SXCIWgW518fqEVBRGvCESy2FDowR79btz/YpKEE6S2G1GlfFlMLCoaULqbVDp9FWUSDUef5seS7eG9AlTHXVQN8mMD+hZmoD5o/og39dkWtSbRLXprpTwP2ZkFTefUjTkGZ0+NOIeMJAnZs+8= Received: by 10.70.131.19 with SMTP id e19mr760132wxd; Wed, 09 Nov 2005 05:17:35 -0800 (PST) Received: by 10.70.9.10 with HTTP; Wed, 9 Nov 2005 05:17:35 -0800 (PST) Message-ID: <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> Date: Wed, 9 Nov 2005 13:17:35 +0000 From: Joao Barros To: Jeremie Le Hen In-Reply-To: <20051109060931.GD5188@obiwan.tataz.chchile.org> 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> Cc: freebsd-performance@freebsd.org 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: Fri, 11 Nov 2005 04:21:12 -0000 On 11/9/05, Jeremie Le Hen wrote: > Hi, Joao, > > > Last month I started a thread[1] on current@ about this, but I guess I > > should have done it here, my apologies for that. > > > > After my initial post I did some more testing and I'm going to start > > clean here with all my findings :) > > > > I started with Samba 3 installed on a PIII 733MHz with fxp (82559) and > > a RAID5 consisting of 4 drives connected to an amr. > > Performance reading or writing was poor, around 5.5MB/s measured on > > two Windows clients and iostat never topped that by much. > > cpu was mbufs were available and there were no IRQs shared. > > To dismiss the amr out of the question I tried with a local IDE > > attached yielding the same results. > > I then tested the same on a machine I have at work, an HP Proliant > > server, Pentium 4 3.06GHz, used SMP instead of GENERIC to use HTT. > > I could get 8MB/s with 2 read or write simultaneous operations. With 1 > > operation I still can only get 6MB/s > > This machine has 1GB ram and after copying a 700MB file to it it was > > all cached. > > A copy to dev/null took 1 second. > > A copy via samba took the same time as if there was no cache for it. > > iostat always showed 0.0 during the operation so that pretty much > > takes disks, controllers, IO out of the picture. > > > > Both machines have cpu, IO and mbufs to spare and they still can't use > > them. Why? > > I won't be able to help you much, but as almost nobody answered you, > I take it for the moment in order to ask you some more informations. > > Which scheduler are you using, 4BSD or ULE ? It might be worth testing > the other one and sending us the new benchmark results. The testings were all with either GENERIC or SMP thus using 4BSD, I can try ULE and see if I get any different results. > > Also, if you are able to remove a drive from your RAID5, you can try > R/W performances from/to it, without using amr(4), both with 4BSD and > ULE. I tried using a single drive, an IDE and a SCSI-2 and on 2 machines at work both with a RAID1. Even better, there is a part in my initial email where I mention that having a 700MB file cached (iostat reported no reads) the results were the same. With this in mind I don't think the problem is at the storage level. -- Joao Barros From owner-freebsd-performance@FreeBSD.ORG Fri Nov 11 12:13:44 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 0326916A41F for ; Fri, 11 Nov 2005 12:13:44 +0000 (GMT) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [209.31.154.42]) by mx1.FreeBSD.org (Postfix) with ESMTP id ACC3D43D45 for ; Fri, 11 Nov 2005 12:13:43 +0000 (GMT) (envelope-from rwatson@FreeBSD.org) Received: from fledge.watson.org (fledge.watson.org [209.31.154.41]) by cyrus.watson.org (Postfix) with ESMTP id BC30D46C24; Fri, 11 Nov 2005 07:13:40 -0500 (EST) Date: Fri, 11 Nov 2005 12:13:40 +0000 (GMT) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Joao Barros In-Reply-To: <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> Message-ID: <20051111121144.P68007@fledge.watson.org> References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> <20051109060931.GD5188@obiwan.tataz.chchile.org> <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: freebsd-performance@freebsd.org, Jeremie Le Hen 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: Fri, 11 Nov 2005 12:13:44 -0000 On Wed, 9 Nov 2005, Joao Barros wrote: > I tried using a single drive, an IDE and a SCSI-2 and on 2 machines at > work both with a RAID1. Even better, there is a part in my initial email > where I mention that having a 700MB file cached (iostat reported no > reads) the results were the same. With this in mind I don't think the > problem is at the storage level. My feeling is that it is likely related to the network stack -- perhaps a combination of Samba<->sockets and TCP<->Windows, hence pointing at things like TCP settings. It would be quite interesting to know what top -S looks like during (perhaps two minutes into) a long-haul transfer, and the output of systat -vmstat 1. I know Apple has done some work on the efficiency of the FreeBSD SMB client when talking to windows servers relating to how TCP behaves, but haven't heard anything about them looking at similar issues in Samba. Robert N M Watson From owner-freebsd-performance@FreeBSD.ORG Fri Nov 11 13:55:04 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 3C85916A41F for ; Fri, 11 Nov 2005 13:55:04 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.203]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4581143D73 for ; Fri, 11 Nov 2005 13:54:50 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so547580wxc for ; Fri, 11 Nov 2005 05:54:49 -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=odw1SU2ZTi12CFuK8EVpPGQuhVL5+tlVorNsynX60FartXY2tWd99KAEBTWPSUrBZoXiAmDVkW0chtQlOh2oyR8M8+OOq4ftAN7Gks+5oonanqzCEj0kMd4Psq47znIx9aEngcqH4QEs91PLGLpzNn+5KdUUA2m1VWQ/tjtWGoc= Received: by 10.70.35.9 with SMTP id i9mr2155001wxi; Fri, 11 Nov 2005 05:54:48 -0800 (PST) Received: by 10.70.9.10 with HTTP; Fri, 11 Nov 2005 05:54:48 -0800 (PST) Message-ID: <70e8236f0511110554v77ea474bvdf6a132eef23f954@mail.gmail.com> Date: Fri, 11 Nov 2005 13:54:48 +0000 From: Joao Barros To: Robert Watson In-Reply-To: <20051111121144.P68007@fledge.watson.org> 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> <20051111121144.P68007@fledge.watson.org> Cc: freebsd-performance@freebsd.org, Jeremie Le Hen 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: Fri, 11 Nov 2005 13:55:04 -0000 On 11/11/05, Robert Watson wrote: > > On Wed, 9 Nov 2005, Joao Barros wrote: > > > I tried using a single drive, an IDE and a SCSI-2 and on 2 machines at > > work both with a RAID1. Even better, there is a part in my initial emai= l > > where I mention that having a 700MB file cached (iostat reported no > > reads) the results were the same. With this in mind I don't think the > > problem is at the storage level. > > My feeling is that it is likely related to the network stack -- perhaps a > combination of Samba<->sockets and TCP<->Windows, hence pointing at thing= s > like TCP settings. It would be quite interesting to know what top -S > looks like during (perhaps two minutes into) a long-haul transfer, and th= e > output of systat -vmstat 1. I know Apple has done some work on the > efficiency of the FreeBSD SMB client when talking to windows servers > relating to how TCP behaves, but haven't heard anything about them lookin= g > at similar issues in Samba. > > Robert N M Watson > Output of top -S and systat -vmstat 1 after copying a file twice (read) uncached and then cached: top -S: last pid: 3744; load averages: 0.12, 0.06, 0.02 =20 up 0+23:41:09 13:48:20 100 processes: 3 running, 57 sleeping, 40 waiting CPU states: 0.0% user, 0.0% nice, 2.1% system, 4.9% interrupt, 93.0% id= le Mem: 16M Active, 812M Inact, 179M Wired, 212K Cache, 112M Buf, 116M Free Swap: 2048M Total, 2048M Free PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.71% idle: cpu= 1 12 root 1 171 52 0K 8K RUN 0 23.6H 83.06% idle: cpu= 0 540 j.barros 1 96 0 5736K 2704K select 0 0:22 6.35% smbd 31 root 1 -68 -187 0K 8K WAIT 0 0:22 4.64% irq19: bg= e0 44 root 1 -44 -163 0K 8K WAIT 0 0:13 1.22% swi1: net 45 root 1 -32 -151 0K 8K WAIT 0 0:54 0.00% swi4: clock sio 63 root 1 20 0 0K 8K syncer 0 0:45 0.00% syncer 4 root 1 -8 0 0K 8K - 0 0:05 0.00% g_down 3 root 1 -8 0 0K 8K - 0 0:05 0.00% g_up 47 root 1 -16 0 0K 8K - 0 0:05 0.00% yarrow 2 root 1 -8 0 0K 8K - 0 0:04 0.00% g_event 462 root 1 96 0 3552K 1404K select 0 0:02 0.00% nmbd 68 root 1 12 0 0K 8K - 0 0:02 0.00% schedcpu 425 root 1 96 0 3396K 2236K select 0 0:01 0.00% sendmail 60 root 1 171 52 0K 8K pgzero 0 0:01 0.00% pagezero 49 root 1 -40 -159 0K 8K WAIT 0 0:01 0.00% swi2: cam= bio 55 root 1 0 0 0K 8K tzpoll 0 0:01 0.00% acpi_ther= mal 33 root 1 -64 -183 0K 8K WAIT 0 0:01 0.00% irq21: ci= ss0 529 j.barros 1 96 0 6076K 2272K select 0 0:01 0.00% sshd 61 root 1 -16 0 0K 8K psleep 0 0:00 0.00% bufdaemon 57 root 1 -8 0 0K 8K - 0 0:00 0.00% fdc0 62 root 1 -4 0 0K 8K vlruwt 0 0:00 0.00% vnlru 441 root 1 8 0 1312K 884K nanslp 0 0:00 0.00% cron 3741 root 1 96 0 2300K 1404K CPU0 0 0:00 0.00% top 526 root 1 4 0 6096K 2216K sbwait 0 0:00 0.00% sshd 295 root 1 96 0 1292K 768K select 0 0:00 0.00% syslogd 58 root 1 -16 0 0K 8K psleep 0 0:00 0.00% pagedaemo= n 376 root 1 96 0 1208K 672K select 0 0:00 0.00% usbd 533 root 1 20 0 3752K 2352K pause 0 0:00 0.00% csh 3719 j.barros 1 96 0 6076K 2412K select 0 0:00 0.00% sshd 3716 root 1 4 0 6096K 2400K sbwait 0 0:00 0.00% sshd 429 smmsp 1 20 0 3296K 1980K pause 0 0:00 0.00% sendmail 466 root 1 96 0 5384K 2308K select 0 0:00 0.00% smbd 3723 root 1 5 0 3732K 2296K ttyin 0 0:00 0.00% csh 1 root 1 8 0 724K 244K wait 0 0:00 0.00% init systat -vmstat 1: 2 users Load 0.13 0.06 0.02 Nov 11 13:48 Mem:KB REAL VIRTUAL VN PAGER SWAP PAGER Tot Share Tot Share Free in out in out Act 15660 2460 58664 3832 119084 count All 1029840 4828 3304436 7592 pages Interrupts Proc:r p d s w Csw Trp Sys Int Sof Flt cow 8970 total 1 30 15030 8 2722 8500 2 22 182924 wire 1: at= kb 16352 act 6: fd= c0 3.7%Sys 3.7%Intr 0.4%User 0.0%Nice 92.2%Idl 831256 inact 13: n= px | | | | | | | | | | 212 cache 14: a= ta =3D=3D++ 118872 free 4955 1= 9: bge daefr 1 21: c= is Namei Name-cache Dir-cache prcfr 2007 cpu0:= time Calls hits % hits % react 2007 cpu1:= time pdwake zfod pdpgs Disks da0 pass0 ofod intrn KB/t 2.00 0.00 %slo-z 114464 buf tps 1 0 tfree 5 dirtybuf MB/s 0.00 0.00 78034 desiredvnodes % busy 2 0 54388 numvnodes 19504 freevnodes dmesg: Copyright (c) 1992-2005 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD 6.0-RELEASE #5: Thu Nov 10 13:57:54 WET 2005 root@bio:/usr/obj/usr/src/sys/SMP Timecounter "i8254" frequency 1193182 Hz quality 0 CPU: Intel(R) Pentium(R) 4 CPU 3.06GHz (3065.82-MHz 686-class CPU) Origin =3D "GenuineIntel" Id =3D 0xf29 Stepping =3D 9 Features=3D0xbfebfbff Features2=3D0x4400> Hyperthreading: 2 logical CPUs real memory =3D 1207934976 (1151 MB) avail memory =3D 1173495808 (1119 MB) ACPI APIC Table: FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs cpu0 (BSP): APIC ID: 0 cpu1 (AP): APIC ID: 1 ioapic0: Changing APIC ID to 2 ioapic1: Changing APIC ID to 3 MADT: Forcing active-low polarity and level trigger for SCI ioapic0 irqs 0-15 on motherboard ioapic1 irqs 16-31 on motherboard npx0: [FAST] npx0: on motherboard npx0: INT 16 interface acpi0: on motherboard acpi0: Power Button (fixed) pci_link0: on acpi0 pci_link1: on acpi0 pci_link2: on acpi0 pci_link3: on acpi0 pci_link4: on acpi0 pci_link5: on acpi0 pci_link6: on acpi0 pci_link7: on acpi0 Timecounter "ACPI-safe" frequency 3579545 Hz quality 1000 acpi_timer0: <32-bit timer at 3.579545MHz> port 0x920-0x923 on acpi0 cpu0: on acpi0 cpu1: on acpi0 pcib0: on acpi0 pci0: on pcib0 atapci0: port 0x2010-0x2017,0x2018-0x201b,0x2020-0x2027,0x2028-0x202b,0x2030-0x203f irq 17 at device 2.0 on pci0 ata2: on atapci0 ata3: on atapci0 pci0: at device 3.0 (no driver attached) pci0: at device 4.0 (no driver attached) bge0: mem 0xf6fd0000-0xf6fdffff irq 19 at device 5.0 on pci0 miibus0: on bge0 brgphy0: on miibus0 brgphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseTX, 1000baseTX-FDX, auto bge0: Ethernet address: 00:12:79:92:0e:e9 bge1: mem 0xf6fc0000-0xf6fcffff irq 20 at device 6.0 on pci0 miibus1: on bge1 brgphy1: on miibus1 brgphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseTX, 1000baseTX-FDX, auto bge1: Ethernet address: 00:12:79:92:0e:e0 ciss0: port 0x2800-0x28ff mem 0xf6fb0000-0xf6fb1fff,0xf6f40000-0xf6f7ffff irq 21 at device 7.0 on pci0 ciss0: [GIANT-LOCKED] isab0: at device 15.0 on pci0 isa0: on isab0 atapci1: port 0x1f0-0x1f7,0x3f6,0x170-0x177,0x376,0x2000-0x200f at device 15.1 on pci0 ata0: on atapci1 ata1: on atapci1 ohci0: mem 0xf6f30000-0xf6f30fff irq 15 at device 15.2 on pci0 ohci0: [GIANT-LOCKED] usb0: OHCI version 1.0, legacy support usb0: SMM does not respond, resetting usb0: on ohci0 usb0: USB revision 1.0 uhub0: (0x1166) OHCI root hub, class 9/0, rev 1.00/1.00, addr 1 uhub0: 4 ports with 4 removable, self powered acpi_button0: on acpi0 acpi_tz0: on acpi0 atkbdc0: port 0x60,0x64 irq 1 on acpi0 atkbd0: irq 1 on atkbdc0 kbd0 at atkbd0 atkbd0: [GIANT-LOCKED] sio0: port 0x3f8-0x3ff irq 4 flags 0x10 on acpi0 sio0: type 16550A fdc0: port 0x3f2-0x3f5 irq 6 drq 2 on acpi0 fdc0: [FAST] fd0: <1440-KB 3.5" drive> on fdc0 drive 0 pmtimer0 on isa0 orm0: at iomem 0xc0000-0xc7fff,0xc8000-0xcbfff,0xcc000-0xcd7ff,0xee000-0xeffff on isa0 ppc0: parallel port not found. sc0: at flags 0x100 on isa0 sc0: VGA <16 virtual consoles, flags=3D0x300> sio1: configured irq 3 not in bitmap of probed irqs 0 sio1: port may not be enabled vga0: at port 0x3c0-0x3df iomem 0xa0000-0xbffff on isa0 Timecounters tick every 1.000 msec acd0: CDROM at ata0-master PIO4 da0 at ciss0 bus 0 target 0 lun 0 da0: Fixed Direct Access SCSI-0 device da0: 135.168MB/s transfers da0: 34998MB (71677440 512 byte sectors: 255H 32S/T 8784C) SMP: AP CPU #1 Launched! Trying to mount root from ufs:/dev/da0s1a bge0: link state changed to UP Offtopic: Robert notice the 2000+ interrupts/s per cpu on systat. Is that normal? -- Joao Barros From owner-freebsd-performance@FreeBSD.ORG Fri Nov 11 15:24:44 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 C69A416A420 for ; Fri, 11 Nov 2005 15:24:44 +0000 (GMT) (envelope-from mike@sentex.net) Received: from smarthost1.sentex.ca (smarthost1.sentex.ca [64.7.153.18]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3DB2743D62 for ; Fri, 11 Nov 2005 15:24:41 +0000 (GMT) (envelope-from mike@sentex.net) Received: from pumice6.sentex.ca (pumice6.sentex.ca [64.7.153.21]) by smarthost1.sentex.ca (8.13.4/8.13.4) with ESMTP id jABFOepv006825 for ; Fri, 11 Nov 2005 10:24:40 -0500 (EST) (envelope-from mike@sentex.net) Received: from lava.sentex.ca (pyroxene.sentex.ca [199.212.134.18]) by pumice6.sentex.ca (8.13.4/8.13.4) with ESMTP id jABFOaRa088700; Fri, 11 Nov 2005 10:24:36 -0500 (EST) (envelope-from mike@sentex.net) Received: from simian.sentex.net (simeon.sentex.ca [192.168.43.27]) by lava.sentex.ca (8.13.3/8.13.3) with ESMTP id jABFORHU018544 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 11 Nov 2005 10:24:28 -0500 (EST) (envelope-from mike@sentex.net) Message-Id: <6.2.3.4.0.20051111102241.0833a2c8@64.7.153.2> X-Mailer: QUALCOMM Windows Eudora Version 6.2.3.4 Date: Fri, 11 Nov 2005 10:23:52 -0500 To: Joao Barros From: Mike Tancsa In-Reply-To: <70e8236f0511110554v77ea474bvdf6a132eef23f954@mail.gmail.co m> References: <70e8236f0511050457s5ce6d8batf805fbc9edd91360@mail.gmail.com> <20051109060931.GD5188@obiwan.tataz.chchile.org> <70e8236f0511090517g29b3f887x1b97ef5dec04548@mail.gmail.com> <20051111121144.P68007@fledge.watson.org> <70e8236f0511110554v77ea474bvdf6a132eef23f954@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed X-Virus-Scanned: by amavisd-new X-Scanned-By: MIMEDefang 2.51 on 64.7.153.18 X-Scanned-By: MIMEDefang 2.53 on 64.7.153.21 Cc: freebsd-performance@freebsd.org 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: Fri, 11 Nov 2005 15:24:44 -0000 At 08:54 AM 11/11/2005, Joao Barros wrote: >Copyright (c) 1992-2005 The FreeBSD Project. >Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 > The Regents of the University of California. All rights reserved. >FreeBSD 6.0-RELEASE #5: Thu Nov 10 13:57:54 WET 2005 > root@bio:/usr/obj/usr/src/sys/SMP >Timecounter "i8254" frequency 1193182 Hz quality 0 >CPU: Intel(R) Pentium(R) 4 CPU 3.06GHz (3065.82-MHz 686-class CPU) > Origin = "GenuineIntel" Id = 0xf29 Stepping = 9 > >Features=0xbfebfbff > Features2=0x4400> > Hyperthreading: 2 logical CPUs Do you actually see a performance improvement with HTT ? I have yet to find an app that we use which actually sees a benefit from it. ---Mike From owner-freebsd-performance@FreeBSD.ORG Fri Nov 11 15:45:27 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 2BF5B16A420 for ; Fri, 11 Nov 2005 15:45:27 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.192]) by mx1.FreeBSD.org (Postfix) with ESMTP id F3B0443D6A for ; Fri, 11 Nov 2005 15:45:23 +0000 (GMT) (envelope-from joao.barros@gmail.com) Received: by xproxy.gmail.com with SMTP id t12so573194wxc for ; Fri, 11 Nov 2005 07:45:22 -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=KgVy1GE0f5fmIPX78uUaii9o7kDIsA81DGxCEPtLB3EBoVM0ixqk/E/kTYno2r0mcnq2/Kds94xfA1C4DO4oLu/SAHzfjxjL3acSy4EAmrnZg17M6nqj+4p8/UD7io6l3x7ei66kCe96ZQGfOFbI1i4M3BbKz26AzrG/TRgXlFg= Received: by 10.70.9.3 with SMTP id 3mr2185982wxi; Fri, 11 Nov 2005 07:45:20 -0800 (PST) Received: by 10.70.9.10 with HTTP; Fri, 11 Nov 2005 07:45:19 -0800 (PST) Message-ID: <70e8236f0511110745w475ff89bu22acd571fd2f4618@mail.gmail.com> Date: Fri, 11 Nov 2005 15:45:19 +0000 From: Joao Barros To: Mike Tancsa In-Reply-To: <6.2.3.4.0.20051111102241.0833a2c8@64.7.153.2> 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> <20051111121144.P68007@fledge.watson.org> <70e8236f0511110554v77ea474bvdf6a132eef23f954@mail.gmail.com> <6.2.3.4.0.20051111102241.0833a2c8@64.7.153.2> Cc: freebsd-performance@freebsd.org 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: Fri, 11 Nov 2005 15:45:27 -0000 On 11/11/05, Mike Tancsa wrote: > At 08:54 AM 11/11/2005, Joao Barros wrote: > >Copyright (c) 1992-2005 The FreeBSD Project. > >Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 > > The Regents of the University of California. All rights reserve= d. > >FreeBSD 6.0-RELEASE #5: Thu Nov 10 13:57:54 WET 2005 > > root@bio:/usr/obj/usr/src/sys/SMP > >Timecounter "i8254" frequency 1193182 Hz quality 0 > >CPU: Intel(R) Pentium(R) 4 CPU 3.06GHz (3065.82-MHz 686-class CPU) > > Origin =3D "GenuineIntel" Id =3D 0xf29 Stepping =3D 9 > > > >Features=3D0xbfebfbff > > Features2=3D0x4400> > > Hyperthreading: 2 logical CPUs > > Do you actually see a performance improvement with HTT ? I have yet > to find an app that we use which actually sees a benefit from it. > > ---Mike The reason behind enabling HTT was not performance. I'm using this machine to port bio and I wanted a SMP like enviroment to test as I only have a UP system at home. -- Joao Barros From owner-freebsd-performance@FreeBSD.ORG Fri Nov 11 20:54:52 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 D916216A41F; Fri, 11 Nov 2005 20:54:52 +0000 (GMT) (envelope-from Alexander@Leidinger.net) Received: from www.ebusiness-leidinger.de (jojo.ms-net.de [84.16.236.246]) by mx1.FreeBSD.org (Postfix) with ESMTP id D243643D9D; Fri, 11 Nov 2005 20:54:32 +0000 (GMT) (envelope-from Alexander@Leidinger.net) Received: from Andro-Beta.Leidinger.net (p54A5E21A.dip.t-dialin.net [84.165.226.26]) (authenticated bits=0) by www.ebusiness-leidinger.de (8.13.1/8.13.1) with ESMTP id jABKVbp2012719; Fri, 11 Nov 2005 21:31:38 +0100 (CET) (envelope-from Alexander@Leidinger.net) Received: from Magellan.Leidinger.net (Magellan.Leidinger.net [192.168.1.1]) by Andro-Beta.Leidinger.net (8.13.3/8.13.3) with ESMTP id jABKsPin092576; Fri, 11 Nov 2005 21:54:25 +0100 (CET) (envelope-from Alexander@Leidinger.net) Date: Fri, 11 Nov 2005 21:54:24 +0100 From: Alexander Leidinger To: Joao Barros Message-ID: <20051111215424.1f553135@Magellan.Leidinger.net> In-Reply-To: <70e8236f0511100231k67d8c940p2a1875d0c09f406d@mail.gmail.com> 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> X-Mailer: Sylpheed-Claws 1.9.100 (GTK+ 2.8.6; i386-portbld-freebsd7.0) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new Cc: Mike Jakubik , freebsd-performance@freebsd.org, Robert Watson , 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: Fri, 11 Nov 2005 20:54:53 -0000 On Thu, 10 Nov 2005 10:31:03 +0000 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 identical > > > 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 thats > > > 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 does > > > not seem to be any visible bottlenecks, its just simply slow. It would > > > 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 Please try if http://leaf.dragonflybsd.org/mailarchive/kernel/2005-07/msg00036.html applies to FreeBSD and test it with inflight enabled. Bye, Alexander. -- Yes, I've heard of "decaf." What's your point? http://www.Leidinger.net Alexander @ Leidinger.net GPG fingerprint = C518 BC70 E67F 143F BE91 3365 79E2 9C60 B006 3FE7