From owner-freebsd-performance@FreeBSD.ORG Sun Sep 24 07:36:31 2006 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 49E9216A403 for ; Sun, 24 Sep 2006 07:36:31 +0000 (UTC) (envelope-from mihai@duras.ro) Received: from mail.duras.ro (mail.duras.ro [86.105.56.133]) by mx1.FreeBSD.org (Postfix) with ESMTP id CAE3443D55 for ; Sun, 24 Sep 2006 07:36:30 +0000 (GMT) (envelope-from mihai@duras.ro) Received: from localhost (localhost [127.0.0.1]) by mail.duras.ro (Postfix) with ESMTP id 64DAF1873F3 for ; Sun, 24 Sep 2006 10:36:31 +0300 (EEST) Received: from mail.duras.ro ([127.0.0.1]) by localhost (mail [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 18243-07 for ; Sun, 24 Sep 2006 10:36:30 +0300 (EEST) Received: from [86.105.56.194] (ma.plimb.cu.barca.prin.padure.ro [86.105.56.194]) by mail.duras.ro (Postfix) with ESMTP id 72838187285 for ; Sun, 24 Sep 2006 10:36:30 +0300 (EEST) Message-ID: <45163558.30905@duras.ro> Date: Sun, 24 Sep 2006 10:35:52 +0300 From: Mihai Tanasescu User-Agent: Mozilla Thunderbird 1.0.6 (Windows/20050716) X-Accept-Language: en-us, en MIME-Version: 1.0 To: freebsd-performance@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new-20030616-p10 (RedHat) at duras.ro Subject: Routing performance HP-ML350 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: Sun, 24 Sep 2006 07:36:31 -0000 Hello, I've got a HP ML-350 machine, dual Xeon 3Ghz. The system is running: FreeBSD 6.1-RELEASE-p7 and has the following netcards and traffic: 2 x fxp cards 2 x em cards each fxp card does about 80mbits traffic in/out each em card does about 200 mbits traffic in/out (packets are routed from em0 to em1 --> cisco 3750 access layer equipment) What I've noticed recently is that my CPU usage stays at about 80% when systat -ip 10 shows an average of 150-180 kpps being forwarded through this box. With SMP enabled I had 85-90% CPU interrupt load so I disabled that. Also I disabled the ipfw queueing code I had on one of the fxp cards (dynamic queueing with 6000 hash buckets - I have quite that many ip addresses and wanted each to hash to a different queue). The load dropped to 60%. Also I have polling disabled on the em cards (polling was causing packet drops with different Hz values; didn't try the idle poll as that used to crash my machine after a day or so of working a couple of months ago) and fast_forwarding enabled. On the fxp cards I tried with both polling enabled/disabled. I wanted to ask if there is any other tweaking possible to improve the routing performance. Thanks, Mihai From owner-freebsd-performance@FreeBSD.ORG Sun Sep 24 14:05:09 2006 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 1837C16A492 for ; Sun, 24 Sep 2006 14:05:08 +0000 (UTC) (envelope-from lists@wm-access.no) Received: from lakepoint.domeneshop.no (lakepoint.domeneshop.no [194.63.248.54]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3645843D53 for ; Sun, 24 Sep 2006 14:05:07 +0000 (GMT) (envelope-from lists@wm-access.no) Received: from [192.168.9.8] (gw1.arcticwireless.no [194.19.37.70]) (authenticated bits=0) by lakepoint.domeneshop.no (8.13.8/8.13.8) with ESMTP id k8OE556Z004177 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NOT); Sun, 24 Sep 2006 16:05:06 +0200 Message-ID: <45169091.4040000@wm-access.no> Date: Sun, 24 Sep 2006 16:05:05 +0200 From: =?ISO-8859-1?Q?Sten_Daniel_S=F8rsdal?= User-Agent: Thunderbird 1.5.0.7 (Windows/20060909) MIME-Version: 1.0 To: Jan Zacharias References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Cc: freebsd-performance@freebsd.org Subject: Re: Samba performance, TCP Stack Issue? 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: Sun, 24 Sep 2006 14:05:09 -0000 Jan Zacharias wrote: > Hi, >=20 > was anybody able to tune the Samba service (2.2.12 or 3.0.23b) so that > it would perform out as well as > unter linux where it maxes out at about 11 Mb/s using some Intel/3com c= ard? >=20 > So far, the most I got out of it was ~6 Mb/s while txing one file to a > win2k station, curiously I get > ~8.5 Mb/s when txing multiple files in parallel. >=20 > So far i messed with: >=20 > - ifconfig mtu > - net.inet.tcp settings > - smb's socket options >=20 > Tuning SO_SNDBUF gave only very little gain, adapting the mtu to match > the win2k box was useless. > W/a net.inet.tcp.inflight.enable enabled the txrate is 2 Mb/s lower, > however 6 Mb/s is still SLOW. >=20 > I cant find the bottleneck, as the system is quite idle. >=20 > Any Suggestions? >=20 Try setting sysctl net.inet.tcp.delayed_ack to '0'. If that helps noticably you may tune sysctl net.inet.tcp.delacktime to a lower value than the default ('100' ??). delacktime is only in effect when net.inet.tcp.delayed_ack =3D '1'. Does that help any? --=20 Sten Daniel S=F8rsdal From owner-freebsd-performance@FreeBSD.ORG Tue Sep 26 15:48:05 2006 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 2818E16A4E7 for ; Tue, 26 Sep 2006 15:48:05 +0000 (UTC) (envelope-from mark@gaiahost.coop) Received: from biodiesel.gaiahost.coop (biodiesel.gaiahost.coop [64.95.78.120]) by mx1.FreeBSD.org (Postfix) with ESMTP id 573E343D83 for ; Tue, 26 Sep 2006 15:46:39 +0000 (GMT) (envelope-from mark@gaiahost.coop) Received: from gaiahost.coop (host-64-65-195-19.spr.choiceone.net [::ffff:64.65.195.19]) (AUTH: LOGIN mark@hubcapconsulting.com) by biodiesel.gaiahost.coop with esmtp; Tue, 26 Sep 2006 11:46:37 -0400 id 00794117.45194B5E.00007B0E Received: by gaiahost.coop (sSMTP sendmail emulation); Tue, 26 Sep 2006 11:46:44 -0400 Date: Tue, 26 Sep 2006 11:46:44 -0400 From: Mark Bucciarelli To: freebsd-performance@freebsd.org Message-ID: <20060926154643.GA3064@rabbit> Mail-Followup-To: freebsd-performance@freebsd.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline User-Agent: Mutt/1.4.2.1i Subject: Why are disk writes so slow? 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: Tue, 26 Sep 2006 15:48:05 -0000 I am reading Richard Stevens' "Advanced Programming in the UNIX Environment," a most excellent book. Out of curiosity, I tried his I/O efficiency program on my IBM A30 Thinkpad, running 6.0-RELEASE with default tuning parameters. The test program reads file on stdin and writes to stdout, and you modify bufsize to watch how time changes. As in his example (with a bufsize of 8192), time ./a.out < 1.5M-testfile > /dev/null runs five times faster than (clock time) time ./a.out < 1.5M-testfile > /a.out.out Can someone explain to me why writing is five times as slow as writing? What's going on in the computer? The file is not O_SYNC, so it can't be validating the data on the disk. Later in the same chapter, he shows the impact of O_SYNC flag. I re-ran this experiment too, and while everything is two orders of magnitude faster than his times in the book, the relative speed of writing with O_SYNC is three times slower. 1993 2006 ----- ---- normal write 2.3s .023s O_SYNC 13.4s .364s slowdow factor 5.8 15.8 Is this all b/c disks are so much larger? m From owner-freebsd-performance@FreeBSD.ORG Tue Sep 26 17:03:07 2006 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 8FC4616A49E for ; Tue, 26 Sep 2006 17:03:07 +0000 (UTC) (envelope-from wmoran@collaborativefusion.com) Received: from mx00.pub.collaborativefusion.com (mx00.pub.collaborativefusion.com [206.210.89.199]) by mx1.FreeBSD.org (Postfix) with ESMTP id 11F6D43D55 for ; Tue, 26 Sep 2006 17:03:01 +0000 (GMT) (envelope-from wmoran@collaborativefusion.com) Received: from collaborativefusion.com (mx01.pub.collaborativefusion.com [206.210.89.201]) (TLS: TLSv1/SSLv3,256bits,AES256-SHA) by wingspan with esmtp; Tue, 26 Sep 2006 13:03:01 -0400 id 00056421.45195D45.0000BC46 Received: from Internal Mail-Server (206.210.89.202) by mx01 (envelope-from wmoran@collaborativefusion.com) with AES256-SHA encrypted SMTP; 26 Sep 2006 13:01:08 -0400 Date: Tue, 26 Sep 2006 13:03:00 -0400 From: Bill Moran To: Mark Bucciarelli Message-Id: <20060926130300.3dc4b06a.wmoran@collaborativefusion.com> In-Reply-To: <20060926154643.GA3064@rabbit> References: <20060926154643.GA3064@rabbit> Organization: Collaborative Fusion X-Mailer: Sylpheed version 2.2.7 (GTK+ 2.8.20; i386-portbld-freebsd6.1) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: freebsd-performance@freebsd.org Subject: Re: Why are disk writes so slow? 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: Tue, 26 Sep 2006 17:03:07 -0000 In response to Mark Bucciarelli : > I am reading Richard Stevens' "Advanced Programming in the UNIX > Environment," a most excellent book. > > Out of curiosity, I tried his I/O efficiency program on my IBM > A30 Thinkpad, running 6.0-RELEASE with default tuning parameters. > The test program reads file on stdin and writes to stdout, and > you modify bufsize to watch how time changes. > > As in his example (with a bufsize of 8192), > > time ./a.out < 1.5M-testfile > /dev/null > > runs five times faster than (clock time) > > time ./a.out < 1.5M-testfile > /a.out.out > > Can someone explain to me why writing is five times as slow as > writing? What's going on in the computer? > > The file is not O_SYNC, so it can't be validating the data on the > disk. > > Later in the same chapter, he shows the impact of O_SYNC flag. I > re-ran this experiment too, and while everything is two orders of > magnitude faster than his times in the book, the relative speed > of writing with O_SYNC is three times slower. > > 1993 2006 > ----- ---- > normal write 2.3s .023s > O_SYNC 13.4s .364s > slowdow factor 5.8 15.8 > > Is this all b/c disks are so much larger? I'm rather confused as to exactly what your question is ... First off, writes are slower than reads if the data you're reading is already cached in RAM. Unless you have _very_ little RAM in your machine, then anything that takes .023s to write is going to be able to fit entirely in the buffer cache, thus repeated access doesn't require any real disk activity. Secondly, as to why Stevens saw less of a slowdown with O_SYNC than you did -- I doubt there's one easy answer. Discs are manufactured differently now than they were in 93, and there's even a huge difference between different brands and different types (i.e. SCSI/SATA) in addition to the differences in hardware connecting the disks, and the drivers for that hardware. There are dozens of places where the difference could be occurring. I would guess that the drive itself does write caching, and this heavily optimizes async writes, but can't improve the performance of sync writes any. -- Bill Moran Collaborative Fusion Inc. From owner-freebsd-performance@FreeBSD.ORG Tue Sep 26 18:05:40 2006 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 4067916A412 for ; Tue, 26 Sep 2006 18:05:40 +0000 (UTC) (envelope-from mark@gaiahost.coop) Received: from biodiesel.gaiahost.coop (biodiesel.gaiahost.coop [64.95.78.120]) by mx1.FreeBSD.org (Postfix) with ESMTP id 91DAA43D60 for ; Tue, 26 Sep 2006 18:05:36 +0000 (GMT) (envelope-from mark@gaiahost.coop) Received: from gaiahost.coop (host-64-65-195-19.spr.choiceone.net [::ffff:64.65.195.19]) (AUTH: LOGIN mark@hubcapconsulting.com) by biodiesel.gaiahost.coop with esmtp; Tue, 26 Sep 2006 14:05:34 -0400 id 007BC0F7.45196BEE.0000777D Received: by gaiahost.coop (sSMTP sendmail emulation); Tue, 26 Sep 2006 14:05:38 -0400 Date: Tue, 26 Sep 2006 14:05:38 -0400 From: Mark Bucciarelli To: Bill Moran Message-ID: <20060926180538.GH3064@rabbit> Mail-Followup-To: Bill Moran , freebsd-performance@freebsd.org References: <20060926154643.GA3064@rabbit> <20060926130300.3dc4b06a.wmoran@collaborativefusion.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline In-Reply-To: <20060926130300.3dc4b06a.wmoran@collaborativefusion.com> User-Agent: Mutt/1.4.2.1i Cc: freebsd-performance@freebsd.org Subject: Re: Why are disk writes so slow? 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: Tue, 26 Sep 2006 18:05:40 -0000 On Tue, Sep 26, 2006 at 01:03:00PM -0400, Bill Moran wrote: > In response to Mark Bucciarelli : > > > Can someone explain to me why writing is five times as slow as > > writing? What's going on in the computer? > > I'm rather confused as to exactly what your question is ... > Heh, I'm just trying to understand how my computer works. I was surprised that writes were _so much_ slower than reads. I figured somebody here knew. > First off, writes are slower than reads if the data you're > reading is already cached in RAM. Unless you have _very_ > little RAM in your machine, then anything that takes .023s to > write is going to be able to fit entirely in the buffer cache, > thus repeated access doesn't require any real disk activity. I could try running the test immediately after rebooting. Although I have no idea if he did. I bet he didn't. And we both used the same size file: 1.5M. I wonder how much RAM he had. > Secondly, as to why Stevens saw less of a slowdown with O_SYNC > than you did -- I doubt there's one easy answer. Discs are > manufactured differently now than they were in 93, and there's > even a huge difference between different brands and different > types (i.e. SCSI/SATA) in addition to the differences in > hardware connecting the disks, and the drivers for that > hardware. There are dozens of places where the difference > could be occurring. I would guess that the drive itself does > write caching, and this heavily optimizes async writes, but > can't improve the performance of sync writes any. I see. Yeah, if you look at normal write compared to read speed that did improve between the two data points. So that supports your conjecture. So you think this data has no value? 1993 2006 ----- ---- (1) /dev/null write 0.3s .005s <-- read speed (2) normal write 2.3s .023s (3) O_SYNC 13.4s .364s (2) / (1) 7.6 4.6 1.5x faster (3) / (2) 5.8 15.8 2.75x slower relative to ASYNC (3) / (1) 44.6 72.8 1.5x slower relative to read It does makes me wonder how this test runs on Linux, since I think databases use O_SYNC. I guess I'd have to install Linux on my laptop and run the same test to have any useful information. And reboot between each test. And shut down network interface and all daemons. In any case, thanks for your reply. m From owner-freebsd-performance@FreeBSD.ORG Tue Sep 26 20:33:53 2006 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 DE31B16A407 for ; Tue, 26 Sep 2006 20:33:53 +0000 (UTC) (envelope-from freebsd-performance@dfmm.org) Received: from dfmm.org (treehorn.dfmm.org [66.180.195.213]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6CB8243D5A for ; Tue, 26 Sep 2006 20:33:51 +0000 (GMT) (envelope-from freebsd-performance@dfmm.org) Received: (qmail 36204 invoked by uid 1000); 26 Sep 2006 20:33:51 -0000 Received: from localhost (sendmail-bs@127.0.0.1) by localhost with SMTP; 26 Sep 2006 20:33:51 -0000 Date: Tue, 26 Sep 2006 13:33:50 -0700 (PDT) From: Jason Stone X-X-Sender: jason@treehorn.dfmm.org To: Mark Bucciarelli In-Reply-To: <20060926154643.GA3064@rabbit> Message-ID: <20060926132742.Q2780@treehorn.dfmm.org> References: <20060926154643.GA3064@rabbit> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: freebsd-performance@freebsd.org Subject: Re: Why are disk writes so slow? 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: Tue, 26 Sep 2006 20:33:53 -0000 -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 > As in his example (with a bufsize of 8192), > > time ./a.out < 1.5M-testfile > /dev/null > > runs five times faster than (clock time) > > time ./a.out < 1.5M-testfile > /a.out.out a) your 1.5M-testfile is most likely still in the cache from previous test runs or from when you created it. b) reading and writing to the same disk, you're going to thrash the disk with seeks. so, some other experiments to try might include: a) create a whole bunch of test files, reboot, and then make sure you use a different test file for every run. b) try variations where you use a ramdisk for the read and disk for the write, then a disk for the read and a ramdisk for the write, and then a ramdsik for both. c) try reading from /dev/zero and writing to disk as the converse of reading from disk and writing to /dev/null, etc. -Jason -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.0 (FreeBSD) Comment: See https://private.idealab.com/public/jason/jason.gpg iD8DBQFFGY6vswXMWWtptckRAu1QAKDg1M3AFoDyHX7Zh3pfMz5RO3zyrQCfcQor z78KtLyYIOKzeaAzq5xYLPY= =Xe8O -----END PGP SIGNATURE----- From owner-freebsd-performance@FreeBSD.ORG Wed Sep 27 14:02:34 2006 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 BC6CA16A40F for ; Wed, 27 Sep 2006 14:02:34 +0000 (UTC) (envelope-from olli@lurza.secnetix.de) Received: from lurza.secnetix.de (lurza.secnetix.de [83.120.8.8]) by mx1.FreeBSD.org (Postfix) with ESMTP id E705D43D45 for ; Wed, 27 Sep 2006 14:02:33 +0000 (GMT) (envelope-from olli@lurza.secnetix.de) Received: from lurza.secnetix.de (dybkve@localhost [127.0.0.1]) by lurza.secnetix.de (8.13.4/8.13.4) with ESMTP id k8RE2R6s019420; Wed, 27 Sep 2006 16:02:32 +0200 (CEST) (envelope-from oliver.fromme@secnetix.de) Received: (from olli@localhost) by lurza.secnetix.de (8.13.4/8.13.1/Submit) id k8RE2RTZ019419; Wed, 27 Sep 2006 16:02:27 +0200 (CEST) (envelope-from olli) Date: Wed, 27 Sep 2006 16:02:27 +0200 (CEST) Message-Id: <200609271402.k8RE2RTZ019419@lurza.secnetix.de> From: Oliver Fromme To: freebsd-performance@FreeBSD.ORG, mark@gaiahost.coop In-Reply-To: <20060926180538.GH3064@rabbit> X-Newsgroups: list.freebsd-performance User-Agent: tin/1.8.0-20051224 ("Ronay") (UNIX) (FreeBSD/4.11-STABLE (i386)) X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-2.1.2 (lurza.secnetix.de [127.0.0.1]); Wed, 27 Sep 2006 16:02:32 +0200 (CEST) Cc: Subject: Re: Why are disk writes so slow? X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: freebsd-performance@FreeBSD.ORG, mark@gaiahost.coop List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 27 Sep 2006 14:02:34 -0000 Mark Bucciarelli wrote: > 1993 2006 > ----- ---- > (1) /dev/null write 0.3s .005s <-- read speed If you mean to say that 1.5 MB have been read in 0.005s, then that's certainly _not_ the read speed of your disk drive. No single drive currently in existence can deliver 300 MB per second. Those 1.5 MB came from the cache. > (2) normal write 2.3s .023s Looks reasonable. > (3) O_SYNC 13.4s .364s Also looks reasonable. Of course it depends a lot on the type of disk (SCSI, ATA), interface speed (PIO*, UDMA*), drive configuration (write caching etc.), vendor of disk and controller, etc. > It does makes me wonder how this test runs on Linux, since I > think databases use O_SYNC. Usually databases issue an fsync() call at important points in time (e.g. after a full transaction was comitted). The performance is better than running all writes synchronously. > I guess I'd have to install Linux on > my laptop and run the same test to have any useful information. What exactly do you try to find out? Linux has different file systems, different virtual memory management, different buffer cache implementation, different scheduler, different I/O drivers ... The numbers that you'll get won't be very useful for comparisons. > And reboot between each test. And shut down network interface > and all daemons. And don't read and write at the same time on the same drive because the disk heads' seek times will blow the benchmark up. If you want to measure write speed, don't read from the disk at the same time, and vice versa. You should use a disk which isn't used for anything else, i.e. don't use the system disk for benchmarks. If you want to benchmark the pure speed of the drive (not the speed of the file system), then don't put a filesystem on the disk at all. Instead, use the raw device. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing Dienstleistungen mit Schwerpunkt FreeBSD: http://www.secnetix.de/bsd Any opinions expressed in this message may be personal to the author and may not necessarily reflect the opinions of secnetix in any way. "If you think C++ is not overly complicated, just what is a protected abstract virtual base pure virtual private destructor, and when was the last time you needed one?" -- Tom Cargil, C++ Journal From owner-freebsd-performance@FreeBSD.ORG Wed Sep 27 16:49:30 2006 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 D5C5D16A47B for ; Wed, 27 Sep 2006 16:49:30 +0000 (UTC) (envelope-from anderson@centtech.com) Received: from mh2.centtech.com (moat3.centtech.com [64.129.166.50]) by mx1.FreeBSD.org (Postfix) with ESMTP id DF23243D45 for ; Wed, 27 Sep 2006 16:49:29 +0000 (GMT) (envelope-from anderson@centtech.com) Received: from [10.177.171.220] (neutrino.centtech.com [10.177.171.220]) by mh2.centtech.com (8.13.1/8.13.1) with ESMTP id k8QGdUUY015495 for ; Tue, 26 Sep 2006 11:39:30 -0500 (CDT) (envelope-from anderson@centtech.com) Message-ID: <451957C2.90701@centtech.com> Date: Tue, 26 Sep 2006 11:39:30 -0500 From: Eric Anderson User-Agent: Thunderbird 1.5.0.7 (X11/20060923) MIME-Version: 1.0 To: freebsd-performance@freebsd.org References: <20060926154643.GA3064@rabbit> In-Reply-To: <20060926154643.GA3064@rabbit> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: ClamAV 0.87.1/1946/Tue Sep 26 08:18:37 2006 on mh2.centtech.com X-Virus-Status: Clean Subject: Re: Why are disk writes so slow? 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, 27 Sep 2006 16:49:30 -0000 On 09/26/06 10:46, Mark Bucciarelli wrote: > I am reading Richard Stevens' "Advanced Programming in the UNIX > Environment," a most excellent book. > > Out of curiosity, I tried his I/O efficiency program on my IBM > A30 Thinkpad, running 6.0-RELEASE with default tuning parameters. > The test program reads file on stdin and writes to stdout, and > you modify bufsize to watch how time changes. > > As in his example (with a bufsize of 8192), > > time ./a.out < 1.5M-testfile > /dev/null > > runs five times faster than (clock time) > > time ./a.out < 1.5M-testfile > /a.out.out > > Can someone explain to me why writing is five times as slow as > writing? What's going on in the computer? > > The file is not O_SYNC, so it can't be validating the data on the > disk. > > Later in the same chapter, he shows the impact of O_SYNC flag. I > re-ran this experiment too, and while everything is two orders of > magnitude faster than his times in the book, the relative speed > of writing with O_SYNC is three times slower. > > 1993 2006 > ----- ---- > normal write 2.3s .023s > O_SYNC 13.4s .364s > slowdow factor 5.8 15.8 > > Is this all b/c disks are so much larger? It's probably because of caching on the disk. The normal write goes in/out of the on-disk cache, the O_SYNC may be forced to go to the platters. Also, if you didn't already, you should run the test many times, umounting/mounting the filesystem in question in between each test. Also, I recommend using a block device, instead of a file on a filesystem, since the filesystem could allocate blocks for the file differently each time, causing varying results. Eric -- ------------------------------------------------------------------------ Eric Anderson Sr. Systems Administrator Centaur Technology Anything that works is better than anything that doesn't. ------------------------------------------------------------------------ From owner-freebsd-performance@FreeBSD.ORG Fri Sep 29 14:21:22 2006 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 0825E16A40F for ; Fri, 29 Sep 2006 14:21:22 +0000 (UTC) (envelope-from cochard@gmail.com) Received: from py-out-1112.google.com (py-out-1112.google.com [64.233.166.179]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1133143D70 for ; Fri, 29 Sep 2006 14:21:20 +0000 (GMT) (envelope-from cochard@gmail.com) Received: by py-out-1112.google.com with SMTP id o67so1201308pye for ; Fri, 29 Sep 2006 07:21:20 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:sender:to:subject:mime-version:content-type:x-google-sender-auth; b=NCzbgrY7fEEqtqeb4fX9eZ/ONTmengD+e9ugk40GUXIwA3C0Upwt+6pRPzNYykX38BHftyr/cAw76eUX+tFLWhEVTAoVbMUwl35dCFuAtFWpvR3L6LJtikg2VU/DKPXxPr+jWdEfAun+xzRUsfkFEQdow5DfuKc2JYNmO5HMrxo= Received: by 10.35.21.1 with SMTP id y1mr742413pyi; Fri, 29 Sep 2006 07:21:20 -0700 (PDT) Received: by 10.35.100.20 with HTTP; Fri, 29 Sep 2006 07:21:20 -0700 (PDT) Message-ID: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> Date: Fri, 29 Sep 2006 16:21:20 +0200 From: "Olivier Cochard-Labbe" Sender: cochard@gmail.com To: freebsd-performance@freebsd.org MIME-Version: 1.0 X-Google-Sender-Auth: fe0f99ca02275405 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable Content-Disposition: inline X-Content-Filtered-By: Mailman/MimeDel 2.1.5 Subject: Samba Performance problem 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, 29 Sep 2006 14:21:22 -0000 Hi all, I meet a performance problem with my customized distro of FreeBSD 6.1(FreeN= AS). Lot's of FreeNAS users compare the performance of FreeNAS with Linux, and the Samba performance are very poor under FreeNAS (but not with NFS or FTP)= . I've use this value on the default samba configuration: socket options =3D IPTOS_LOWDELAY TCP_NODELAY SO_SNDBUF=3D16384 SO_RCVBUF=3D16384 But there is no change. Then I've try to add a "tune" parameter that set this kernel variables: net.inet.tcp.delayed_ack=3D0 net.inet.tcp.sendspace=3D65536 net.inet.tcp.recvspace=3D65536 net.inet.udp.recvspace=3D65536 net.inet.udp.maxdgram=3D57344 net.local.stream.recvspace=3D65535 net.local.stream.sendspace=3D65535 kern.ipc.maxsockbuf=3D2097152 kern.ipc.somaxconn=3D8192 kern.ipc.nmbclusters=3D60000 kern.maxfiles=3D65536 kern.maxfilesperproc=3D32768 net.inet.tcp.inflight.enable=3D0 But there is no change too... How can I improve the Samba performace ? Thanks, Olivier --=20 Olivier Cochard-Labb=E9 FreeNAS main developer http://www.freenas.org Skype: callto://ocochard From owner-freebsd-performance@FreeBSD.ORG Fri Sep 29 16:13:09 2006 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 6EA3116A492 for ; Fri, 29 Sep 2006 16:13:09 +0000 (UTC) (envelope-from mikej@rogers.com) Received: from smtp107.rog.mail.re2.yahoo.com (smtp107.rog.mail.re2.yahoo.com [68.142.225.205]) by mx1.FreeBSD.org (Postfix) with SMTP id 09F5743D7C for ; Fri, 29 Sep 2006 16:13:07 +0000 (GMT) (envelope-from mikej@rogers.com) Received: (qmail 35736 invoked from network); 29 Sep 2006 16:13:07 -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=Z0vKeYjdVuNQ26V0V6GnVA0GANftg31Tgf4TgBQHtnqaUpdwNwXd4dbJOhz0DEHelhTzAUcXvZcXTK9oKK//FsVMLMb5uvwY/S4HeXLFt5VUdomKOaxY4oEIeF3lQkAabMRGvTVUSrwWI/6UXxcUIE0Zy+dESA/ncXZLU4JSsHU= ; Received: from unknown (HELO ?127.0.0.1?) (mikej@rogers.com@74.111.253.239 with plain) by smtp107.rog.mail.re2.yahoo.com with SMTP; 29 Sep 2006 16:13:06 -0000 Message-ID: <451D4630.7040902@rogers.com> Date: Fri, 29 Sep 2006 12:13:36 -0400 From: Mike Jakubik User-Agent: Thunderbird 1.5.0.7 (Windows/20060909) MIME-Version: 1.0 To: Olivier Cochard-Labbe References: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> In-Reply-To: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-performance@freebsd.org Subject: Re: Samba Performance problem 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, 29 Sep 2006 16:13:09 -0000 Olivier Cochard-Labbe wrote: > Hi all, > > I meet a performance problem with my customized distro of FreeBSD > 6.1(FreeNAS). > Lot's of FreeNAS users compare the performance of FreeNAS with Linux, and > the Samba performance are very poor under FreeNAS (but not with NFS or > FTP). > > ... > But there is no change too... How can I improve the Samba performace ? Basically, you can't. Samba has always lacked in performance on FreeBSD, it has been discussed many times and AFAIK no solution has been found. You can also try adding these to smb.conf.: use sendfile = yes strict locking = no From owner-freebsd-performance@FreeBSD.ORG Fri Sep 29 16:27:37 2006 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 C1C4716A407 for ; Fri, 29 Sep 2006 16:27:37 +0000 (UTC) (envelope-from fbsd-performance@graphics.cs.uni-sb.de) Received: from uni-sb.de (uni-sb.de [134.96.252.33]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4E1AF43D5A for ; Fri, 29 Sep 2006 16:27:27 +0000 (GMT) (envelope-from fbsd-performance@graphics.cs.uni-sb.de) Received: from mail.cs.uni-sb.de (mail.cs.uni-sb.de [134.96.254.200]) by uni-sb.de (8.13.8/2006081400) with ESMTP id k8TGRPGr000805 for ; Fri, 29 Sep 2006 18:27:26 +0200 (CEST) Received: from scihparg.cs.uni-sb.de (scihparg.cs.uni-sb.de [134.96.249.75]) by mail.cs.uni-sb.de (8.13.8/2006081400) with ESMTP id k8TGRPC7012829 for ; Fri, 29 Sep 2006 18:27:25 +0200 (CEST) Received: from bofh.cs.uni-sb.de ([134.96.249.208]) by scihparg.cs.uni-sb.de with esmtpsa (TLS-1.0:DHE_RSA_AES_256_CBC_SHA:32) (Exim 4.50) id 1GTLCv-0001As-6Z for freebsd-performance@freebsd.org; Fri, 29 Sep 2006 18:27:25 +0200 In-Reply-To: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> References: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> To: Performance/tuning Message-ID: From: "Jan Zacharias" Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-1 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Date: Fri, 29 Sep 2006 18:27:23 +0200 User-Agent: Opera M2/8.51 (Linux, build 1462) X-SA-Exim-Connect-IP: 134.96.249.208 X-SA-Exim-Mail-From: fbsd-performance@graphics.cs.uni-sb.de X-Spam-Checker-Version: SpamAssassin 3.0.3 (2005-04-27) on scihparg.cs.uni-sb.de X-Spam-Level: X-Spam-Status: No, score=-2.8 required=5.0 tests=ALL_TRUSTED autolearn=disabled version=3.0.3 X-SA-Exim-Version: 4.2 (built Thu, 03 Mar 2005 10:44:12 +0100) X-SA-Exim-Scanned: Yes (on scihparg.cs.uni-sb.de) Subject: Re: Samba Performance problem 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, 29 Sep 2006 16:27:37 -0000 Hi Olivier, generally the Samba performance on *BSD just sucks, I wrote to this list with the same problem some weeks ago, no one was able to tune smbd so that it would outperform as well as under linux. The only solution is to migrate your fileserver to a Linux distribution as I did. Performance on the same Hardware: now ~11MB/s using 100Mbit NIC, before ~6MB/s (FreeBSD 6.1). Greetz, Jan On Fri, 29 Sep 2006 16:21:20 +0200, Olivier Cochard-Labbe wrote: > Hi all, > > I meet a performance problem with my customized distro of FreeBSD > 6.1(FreeNAS). > Lot's of FreeNAS users compare the performance of FreeNAS with Linux, and > the Samba performance are very poor under FreeNAS (but not with NFS or > FTP). > > > I've use this value on the default samba configuration: > socket options = IPTOS_LOWDELAY TCP_NODELAY > SO_SNDBUF=16384 > SO_RCVBUF=16384 > > But there is no change. > > Then I've try to add a "tune" parameter that set this kernel variables: > net.inet.tcp.delayed_ack=0 > net.inet.tcp.sendspace=65536 > net.inet.tcp.recvspace=65536 > net.inet.udp.recvspace=65536 > net.inet.udp.maxdgram=57344 > net.local.stream.recvspace=65535 > net.local.stream.sendspace=65535 > kern.ipc.maxsockbuf=2097152 > kern.ipc.somaxconn=8192 > kern.ipc.nmbclusters=60000 > kern.maxfiles=65536 > kern.maxfilesperproc=32768 > net.inet.tcp.inflight.enable=0 > > But there is no change too... How can I improve the Samba performace ? > > Thanks, > > Olivier > -- Co.G.Wheel - Administation Group, Computer Graphics Lab of Saarland University Phone: 0681 302 3841 Telecopy: 0681 302 3843 Web: http://graphics.cs.uni-sb.de From owner-freebsd-performance@FreeBSD.ORG Fri Sep 29 16:55:46 2006 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 8B55116A6E7 for ; Fri, 29 Sep 2006 16:55:46 +0000 (UTC) (envelope-from david@solexine.fr) Received: from smtp19.orange.fr (smtp19.orange.fr [80.12.242.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id AC1E543DD2 for ; Fri, 29 Sep 2006 16:49:05 +0000 (GMT) (envelope-from david@solexine.fr) Received: from smtp-msa-out19.orange.fr (mwinf1917 [172.22.129.117]) by mwinf1905.orange.fr (SMTP Server) with ESMTP id 7A3525C493A0 for ; Fri, 29 Sep 2006 18:45:07 +0200 (CEST) Received: from [192.168.254.21] (AMontpellier-152-1-49-174.w81-251.abo.wanadoo.fr [81.251.191.174]) by mwinf1917.orange.fr (SMTP Server) with ESMTP id B4EF01C000EF for ; Fri, 29 Sep 2006 18:44:47 +0200 (CEST) X-ME-UUID: 20060929164447741.B4EF01C000EF@mwinf1917.orange.fr Message-ID: <451D4E9E.1060704@solexine.fr> Date: Fri, 29 Sep 2006 18:49:34 +0200 From: David Touitou Organization: Solexine User-Agent: Thunderbird 1.5.0.7 (Windows/20060909) MIME-Version: 1.0 To: Performance/tuning References: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit Subject: Re: Samba Performance problem 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, 29 Sep 2006 16:55:46 -0000 Hi all, Jan Zacharias a écrit : > generally the Samba performance on *BSD just sucks, > > The only solution is to migrate your fileserver to a Linux distribution > as I did. Olivier being the main developper/maintainer of FreeNAS, this is not an option 8) (bravo au fait) David. -- A: Yes. >Q: Are you sure? >>A: Because it reverses the logical flow of conversation. >>>Q: Why is top posting annoying in email? From owner-freebsd-performance@FreeBSD.ORG Fri Sep 29 16:56:02 2006 Return-Path: X-Original-To: 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 A687916A53D for ; Fri, 29 Sep 2006 16:56:02 +0000 (UTC) (envelope-from mikej@rogers.com) Received: from smtp100.rog.mail.re2.yahoo.com (smtp100.rog.mail.re2.yahoo.com [206.190.36.78]) by mx1.FreeBSD.org (Postfix) with SMTP id 90C1743EBE for ; Fri, 29 Sep 2006 16:50:47 +0000 (GMT) (envelope-from mikej@rogers.com) Received: (qmail 53772 invoked from network); 29 Sep 2006 16:50:46 -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=S8rmoAxPXfwFss8rQ9wSkKCOx32qBQzdIhDFdSIFZDNU4SQ3VlO1B7HEW/Dq2dNBtXfQaOjGSMt834MMl4iaGR+4lXwGIx55+5fj7sYhKp2uZyEZBVxpPnGfKjc2mbwFhQq7iOp/vz5np876Vd8ETnyEakUs4q4fsiCmbUZbr9E= ; Received: from unknown (HELO ?127.0.0.1?) (mikej@rogers.com@74.111.253.239 with plain) by smtp100.rog.mail.re2.yahoo.com with SMTP; 29 Sep 2006 16:50:46 -0000 Message-ID: <451D4F07.7020108@rogers.com> Date: Fri, 29 Sep 2006 12:51:19 -0400 From: Mike Jakubik User-Agent: Thunderbird 1.5.0.7 (Windows/20060909) MIME-Version: 1.0 To: Jan Zacharias References: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> <451D4630.7040902@rogers.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: performance@freebsd.org Subject: Re: Samba Performance problem 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, 29 Sep 2006 16:56:02 -0000 Jan Zacharias wrote: > Btw. the same problem exists on MAC OSX, as OSX also uses a BSD Kernel > the problem might be, that samba was optimized for the linux kernel > tcp stack. Quite possibly, just as MySQL has been written with Linux primarily in mind. From owner-freebsd-performance@FreeBSD.ORG Fri Sep 29 15:28:48 2006 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 539C116A412 for ; Fri, 29 Sep 2006 15:28:48 +0000 (UTC) (envelope-from jan@graphics.cs.uni-sb.de) Received: from uni-sb.de (uni-sb.de [134.96.252.33]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9C38843D53 for ; Fri, 29 Sep 2006 15:28:32 +0000 (GMT) (envelope-from jan@graphics.cs.uni-sb.de) Received: from mail.cs.uni-sb.de (mail.cs.uni-sb.de [134.96.254.200]) by uni-sb.de (8.13.8/2006081400) with ESMTP id k8TFSUog011658 for ; Fri, 29 Sep 2006 17:28:30 +0200 (CEST) Received: from scihparg.cs.uni-sb.de (scihparg.cs.uni-sb.de [134.96.249.75]) by mail.cs.uni-sb.de (8.13.8/2006081400) with ESMTP id k8TFSUEu007870 for ; Fri, 29 Sep 2006 17:28:30 +0200 (CEST) Received: from bofh.cs.uni-sb.de ([134.96.249.208]) by scihparg.cs.uni-sb.de with esmtpsa (TLS-1.0:DHE_RSA_AES_256_CBC_SHA:32) (Exim 4.50) id 1GTKHu-0000bd-Cn for freebsd-performance@freebsd.org; Fri, 29 Sep 2006 17:28:30 +0200 Date: Fri, 29 Sep 2006 17:28:27 +0200 To: Performance/tuning References: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> From: "Jan Zacharias" Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-1 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-ID: In-Reply-To: <3131aa530609290721o267d55bakff4e801ef4000675@mail.gmail.com> User-Agent: Opera M2/8.51 (Linux, build 1462) X-SA-Exim-Connect-IP: 134.96.249.208 X-SA-Exim-Mail-From: jan@graphics.cs.uni-sb.de X-Spam-Checker-Version: SpamAssassin 3.0.3 (2005-04-27) on scihparg.cs.uni-sb.de X-Spam-Level: X-Spam-Status: No, score=-3.1 required=5.0 tests=ALL_TRUSTED,AWL autolearn=disabled version=3.0.3 X-SA-Exim-Version: 4.2 (built Thu, 03 Mar 2005 10:44:12 +0100) X-SA-Exim-Scanned: Yes (on scihparg.cs.uni-sb.de) X-Mailman-Approved-At: Fri, 29 Sep 2006 23:18:16 +0000 Subject: Re: Samba Performance problem 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, 29 Sep 2006 15:28:48 -0000 Hi Olivier, generally the Samba performance on *BSD just sucks, I wrote to this list with the same problem some weeks ago, no one was able to tune smbd so that it would outperform as well as under linux. The only solution is to migrate your fileserver to a Linux distribution as I did. Performance on the same Hardware: now ~11MB/s using 100Mbit NIC, before ~6MB/s (FreeBSD 6.1). Greetz, Jan On Fri, 29 Sep 2006 16:21:20 +0200, Olivier Cochard-Labbe wrote: > Hi all, > > I meet a performance problem with my customized distro of FreeBSD > 6.1(FreeNAS). > Lot's of FreeNAS users compare the performance of FreeNAS with Linux, and > the Samba performance are very poor under FreeNAS (but not with NFS or > FTP). > > > I've use this value on the default samba configuration: > socket options = IPTOS_LOWDELAY TCP_NODELAY > SO_SNDBUF=16384 > SO_RCVBUF=16384 > > But there is no change. > > Then I've try to add a "tune" parameter that set this kernel variables: > net.inet.tcp.delayed_ack=0 > net.inet.tcp.sendspace=65536 > net.inet.tcp.recvspace=65536 > net.inet.udp.recvspace=65536 > net.inet.udp.maxdgram=57344 > net.local.stream.recvspace=65535 > net.local.stream.sendspace=65535 > kern.ipc.maxsockbuf=2097152 > kern.ipc.somaxconn=8192 > kern.ipc.nmbclusters=60000 > kern.maxfiles=65536 > kern.maxfilesperproc=32768 > net.inet.tcp.inflight.enable=0 > > But there is no change too... How can I improve the Samba performace ? > > Thanks, > > Olivier > -- Co.G.Wheel - Administation Group, Computer Graphics Lab of Saarland University Phone: 0681 302 3841 Telecopy: 0681 302 3843 Web: http://graphics.cs.uni-sb.de