From owner-freebsd-performance@FreeBSD.ORG Sat Jan 24 01:05:51 2004 Return-Path: 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 A98BD16A4CE for ; Sat, 24 Jan 2004 01:05:51 -0800 (PST) Received: from pillage.dreamhost.com (pillage.dreamhost.com [66.33.213.23]) by mx1.FreeBSD.org (Postfix) with ESMTP id B0AFD43D2F for ; Sat, 24 Jan 2004 01:05:50 -0800 (PST) (envelope-from ron@primenumbers.org) Received: from [192.168.101.17] (c-67-161-80-162.client.comcast.net [67.161.80.162]) by pillage.dreamhost.com (Postfix) with ESMTP id 5CC1814969A; Sat, 24 Jan 2004 01:05:50 -0800 (PST) In-Reply-To: References: Mime-Version: 1.0 (Apple Message framework v609) Content-Type: text/plain; charset=US-ASCII; format=flowed Message-Id: <727B1791-4E4C-11D8-926E-000393DED8CE@primenumbers.org> Content-Transfer-Encoding: 7bit From: Ron Smith Date: Sat, 24 Jan 2004 01:05:25 -0800 To: "HarryH" X-Mailer: Apple Mail (2.609) X-Mailman-Approved-At: Tue, 27 Jan 2004 14:30:09 -0800 cc: freebsd-performance@freebsd.org Subject: Re: A good BSD Text Book? X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 24 Jan 2004 09:05:51 -0000 The book that really made everything clear to me (and I compiled my first kernel from it's instructions) was "The Complete FreeBSD" by Greg Lehey. I had the 3rd edition from the FreeBSD PowerPak (www.freebsdmall.com) 4.something. I think it has been updated and includes some 5.x stuff now. Good luck, -Ron On Jan 23, 2004, at 10:08 PM, HarryH wrote: > Hi, > Can anyone recommend a/some really good BSD (4.8) books/manuals, for > not only a BSD beginner, but for someone that will really get into > detail in a short time? What I found at Border's was a real > beginner's book. Or, will I end up with a couple of books? I have > the "Unix in a Nutshell" by O'Reilly but would like to zero in on BSD. > > Thanks, > Harry > _______________________________________________ > 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" > From owner-freebsd-performance@FreeBSD.ORG Sat Jan 24 04:59:22 2004 Return-Path: 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 91C0816A4CE for ; Sat, 24 Jan 2004 04:59:22 -0800 (PST) Received: from munk.nu (mail.munk.nu [213.152.51.194]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5240943D46 for ; Sat, 24 Jan 2004 04:59:21 -0800 (PST) (envelope-from munk@munk.nu) Received: from munk by munk.nu with local (Exim 4.30; FreeBSD) id 1AkNNc-00094j-Jd; Sat, 24 Jan 2004 12:59:16 +0000 Date: Sat, 24 Jan 2004 12:59:16 +0000 From: Jez Hancock To: HarryH Message-ID: <20040124125916.GC34163@users.munk.nu> Mail-Followup-To: HarryH , freebsd-performance@freebsd.org References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.1i Sender: Jez Hancock X-Mailman-Approved-At: Tue, 27 Jan 2004 14:30:09 -0800 cc: freebsd-performance@freebsd.org Subject: Re: A good BSD Text Book? X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 24 Jan 2004 12:59:22 -0000 On Fri, Jan 23, 2004 at 10:08:25PM -0800, HarryH wrote: > Hi, > Can anyone recommend a/some really good BSD (4.8) books/manuals, for > not only a BSD beginner, but for someone that will really get into > detail in a short time? What I found at Border's was a real > beginner's book. Or, will I end up with a couple of books? I have > the "Unix in a Nutshell" by O'Reilly but would like to zero in on BSD. The files installed into /usr/share/doc/ are generally very good and cover various aspects for both beginners and intermediate freebsd users. Other than that, someone posted a good detailed list of BSD sites (not just FreeBSD) recently on the freebsd-questions list: http://marc.theaimsgroup.com/?l=freebsd-questions&m=107419514917946&w=2 you should find a lot of decent BSD specific reading there. Good luck :P -- Jez Hancock - System Administrator / PHP Developer http://munk.nu/ http://jez.hancock-family.com/ - Another FreeBSD Diary http://ipfwstats.sf.net/ - ipfw peruser traffic logging From owner-freebsd-performance@FreeBSD.ORG Sun Jan 25 03:04:29 2004 Return-Path: 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 07ED116A4CE; Sun, 25 Jan 2004 03:04:29 -0800 (PST) Received: from happygiraffe.net (happygiraffe.net [81.6.215.59]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7F00E43D31; Sun, 25 Jan 2004 03:04:27 -0800 (PST) (envelope-from dom@happygiraffe.net) Received: from localhost (localhost.happygiraffe.net [127.0.0.1]) by happygiraffe.net (Postfix) with ESMTP id 02B4AB83E; Sun, 25 Jan 2004 11:04:26 +0000 (GMT) Received: from happygiraffe.net ([127.0.0.1]) by localhost (ppe.happygiraffe.net [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 35528-04-2; Sun, 25 Jan 2004 11:04:25 +0000 (GMT) Received: by happygiraffe.net (Postfix, from userid 1001) id 52A31B80A; Sun, 25 Jan 2004 11:04:25 +0000 (GMT) Date: Sun, 25 Jan 2004 11:04:25 +0000 To: Robert Watson Message-ID: <20040125110425.GA35789@ppe.happygiraffe.net> References: <003c01c3de8d$d569edb0$471b3dd4@dual> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.5.1i From: dom@happygiraffe.net (Dominic Mitchell) X-Virus-Scanned: by amavisd-new at happygiraffe.net X-Mailman-Approved-At: Tue, 27 Jan 2004 14:30:09 -0800 cc: performance@FreeBSD.ORG cc: current@FreeBSD.ORG cc: Willem Jan Withagen Subject: Re: Old SUN NFS performance papers. X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 25 Jan 2004 11:04:29 -0000 On Sat, Jan 24, 2004 at 09:14:51PM -0500, Robert Watson wrote: > I haven't done much benchmarking on NFS lately, but something worth > remembering is that people have spent a lot of time researching and > optimizing TCP for a variety of connection types, whereas the NFS code > basically has a static implementation of RPC backoff and flow control that > hasn't evolved much. TCP is aware of things like the pathwise-mtu to the > server and adapts, whereas UDP just loses packets due to fragmentation, > especially if you are using larger block sizes. Please do post your > discoveries on performance@, and perhaps we could build an NFS performance > tuning section in the FreeBSD Handbook (or if there's not that much > content, add it to the FAQ)? I'm just playing with this... The first thing to note (probably) is to check that you can ping your server with a similiar size packet to the one you're using. I realised that my network isn't as robust as I thought it was very quickly yesterday, when pinging my server with an 8k packet. I was seeing 70% packet loss. The default ping showed no problems at all. The reason I mention it is that I'd been playing with NFS tuning because I had been seeing lockups. But the fault really lies at a lower level than NFS, it appears. -Dom From owner-freebsd-performance@FreeBSD.ORG Sun Jan 25 07:11:57 2004 Return-Path: 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 DE0BB16A4CE for ; Sun, 25 Jan 2004 07:11:57 -0800 (PST) Received: from winston.piwebs.com (217-19-20-186.dsl.cambrium.nl [217.19.20.186]) by mx1.FreeBSD.org (Postfix) with SMTP id 8D0A743D5A for ; Sun, 25 Jan 2004 07:11:49 -0800 (PST) (envelope-from avleeuwen@piwebs.com) Received: (qmail 54108 invoked from network); 25 Jan 2004 15:11:47 -0000 Received: from vincent.piwebs.com (192.168.0.97) by winston.piwebs.com with SMTP; 25 Jan 2004 15:11:47 -0000 From: Arjan van Leeuwen To: freebsd-performance@freebsd.org Date: Sun, 25 Jan 2004 16:11:46 +0100 User-Agent: KMail/1.5.94 References: In-Reply-To: MIME-Version: 1.0 Content-Disposition: inline Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <200401251611.47084.avleeuwen@piwebs.com> X-Mailman-Approved-At: Tue, 27 Jan 2004 14:30:09 -0800 cc: HarryH Subject: Re: A good BSD Text Book? X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: avleeuwen@piwebs.com List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 25 Jan 2004 15:11:58 -0000 On Saturday 24 January 2004 07:08, HarryH wrote: > Hi, > Can anyone recommend a/some really good BSD (4.8) books/manuals, for not > only a BSD beginner, but for someone that will really get into detail in a > short time? What I found at Border's was a real beginner's book. Or, will > I end up with a couple of books? I have the "Unix in a Nutshell" by > O'Reilly but would like to zero in on BSD. I really like Greg Lehey's 'The Complete FreeBSD'. It seems to go deeper into the material than Absolute BSD from Michael Lucas (although I think that's a great book too, it's very easy to read). Other books that I haven't read myself but I heard good things about are 'FreeBSD Unleashed' and 'FreeBSD: The Complete Reference'. Arjan From owner-freebsd-performance@FreeBSD.ORG Mon Jan 26 21:49:13 2004 Return-Path: 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 8C62616AAAA; Mon, 26 Jan 2004 21:49:11 -0800 (PST) Received: from mail.seekingfire.com (coyote.seekingfire.com [24.72.10.212]) by mx1.FreeBSD.org (Postfix) with ESMTP id A720144A2A; Mon, 26 Jan 2004 20:52:09 -0800 (PST) (envelope-from tillman@seekingfire.com) Received: by mail.seekingfire.com (Postfix, from userid 500) id CE5EF63; Mon, 26 Jan 2004 22:52:08 -0600 (CST) Date: Mon, 26 Jan 2004 22:52:08 -0600 From: Tillman Hodgson To: current@freebsd.org Message-ID: <20040127045208.GC83558@seekingfire.com> References: <003c01c3de8d$d569edb0$471b3dd4@dual> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: X-Habeas-SWE-1: winter into spring X-Habeas-SWE-2: brightly anticipated X-Habeas-SWE-3: like Habeas SWE (tm) X-Habeas-SWE-4: Copyright 2002 Habeas (tm) X-Habeas-SWE-5: Sender Warranted Email (SWE) (tm). The sender of this X-Habeas-SWE-6: email in exchange for a license for this Habeas X-Habeas-SWE-7: warrant mark warrants that this is a Habeas Compliant X-Habeas-SWE-8: Message (HCM) and not spam. Please report use of this X-Habeas-SWE-9: mark in spam to . X-GPG-Key-ID: 828AFC7B X-GPG-Fingerprint: 5584 14BA C9EB 1524 0E68 F543 0F0A 7FBC 828A FC7B X-GPG-Key: http://www.seekingfire.com/gpg_key.asc X-Urban-Legend: There is lots of hidden information in headers User-Agent: Mutt/1.5.5.1i cc: performance@freebsd.org Subject: Re: Old SUN NFS performance papers. X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 27 Jan 2004 05:49:13 -0000 On Sat, Jan 24, 2004 at 09:14:51PM -0500, Robert Watson wrote: > I haven't done much benchmarking on NFS lately, but something worth > remembering is that people have spent a lot of time researching and > optimizing TCP for a variety of connection types, whereas the NFS code > basically has a static implementation of RPC backoff and flow control that > hasn't evolved much. TCP is aware of things like the pathwise-mtu to the > server and adapts, whereas UDP just loses packets due to fragmentation, > especially if you are using larger block sizes. Please do post your > discoveries on performance@, and perhaps we could build an NFS performance > tuning section in the FreeBSD Handbook (or if there's not that much > content, add it to the FAQ)? I once spent a great deal of time with bonnie++, gnuplot and LaTeX generating a nice 10-page or so document on NFS tuning between various systems (the server was always a FreeBSD 4-STABLE box of the 4.7 vintage with several vinum mirrors of various speeds). Naturally, I then lost the document when my home drive disintegated a few months later. i was /just/ in the process of moving my home dir to the file server where it would have multiple levels of backup ... Anyway, I'd be willing to do some more testing and writes up the results for the Handbook (when I get back from a much-needed vacation) if I could obtain several collaborators with a variety of hardware (NFS testing takes /much/ time to cover all the variants and hardware possibilities) and a good idea from someone (such as yourself) on what sort of testing would be seen as valuable. For instance, my primary concern is remote Maildir speed and backup up several servers onto a backup host (where it can spooled out to tape via some scripts). So my access patterns are fairly specific, and likely not typical. Guidance on what a typical NFS user looks like would be precious knowledge ;-) -T -- Page 38: Be sure that, in the excitement of creating a totally rad password, you resist the temptation to tell someone just to show off how smart you are. - Harley Hahn, _The Unix Companion_ From owner-freebsd-performance@FreeBSD.ORG Tue Jan 27 11:20:20 2004 Return-Path: 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 83F4316A4D0; Tue, 27 Jan 2004 11:20:20 -0800 (PST) Received: from manian.sics.se (manian.sics.se [193.10.66.13]) by mx1.FreeBSD.org (Postfix) with ESMTP id DD78B43D6E; Tue, 27 Jan 2004 11:19:23 -0800 (PST) (envelope-from bg@sics.se) Received: from sics.se (localhost [127.0.0.1]) by manian.sics.se (8.12.10/8.12.10) with SMTP id i0RJHwBS021033; Tue, 27 Jan 2004 20:17:58 +0100 (CET) (envelope-from bg@sics.se) Date: Tue, 27 Jan 2004 20:17:58 +0100 From: =?ISO-8859-1?Q?Bj=F6rn_Gr=F6nvall?= To: Robert Watson Message-Id: <20040127201758.13be4f08.bg@sics.se> In-Reply-To: References: <003c01c3de8d$d569edb0$471b3dd4@dual> Organization: SICS.SE X-Mailer: Sylpheed version 0.9.6claws (GTK+ 1.2.10; i386-portbld-freebsd5.2) Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="Multipart=_Tue__27_Jan_2004_20_17_58_+0100_=Q2bfUEBOTrwOu39" X-Mailman-Approved-At: Tue, 27 Jan 2004 14:30:09 -0800 cc: bg@sics.se cc: performance@freebsd.org cc: current@freebsd.org cc: Willem Jan Withagen Subject: Re: Old SUN NFS performance papers. X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 27 Jan 2004 19:20:20 -0000 This is a multi-part message in MIME format. --Multipart=_Tue__27_Jan_2004_20_17_58_+0100_=Q2bfUEBOTrwOu39 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit On Sat, 24 Jan 2004 21:14:51 -0500 (EST) Robert Watson wrote: > I haven't done much benchmarking on NFS lately, but something worth > remembering is that people have spent a lot of time researching and > optimizing TCP for a variety of connection types, whereas the NFS code > basically has a static implementation of RPC backoff and flow control that > hasn't evolved much. One reason that FreeBSD users experience poor NFSv3/TCP performance is that the defaults for rsize and wsize is unusually small, only 8k. Solaris and HP-UX defaults to 32k for a good reason. I guess TCP simply needs a little bit more data to chew on to be efficient. I tested this on 5.2-CURRENT and found that large file read performance went up from 56Mbit/s to 80Mbit/s, an improvement by 43%. I have written a patch that makes FreeBSD use the same defaults as Solaris and HP-UX. Note that with NFSv3 there is no risk associated with specifying to large values for [rw]size. The server automatically limits these values in the fsinfo rpc. Patch is attached. Cheers, Bj鰎n -- _ _ ,_______________. Bjorn Gronvall (Bj鰎n Gr鰊vall) /_______________/| Swedish Institute of Computer Science | || PO Box 1263, S-164 29 Kista, Sweden | Schroedingers || Email: bg@sics.se, Phone +46 -8 633 15 25 | Cat |/ Cellular +46 -70 768 06 35, Fax +46 -8 751 72 30 '---------------' --Multipart=_Tue__27_Jan_2004_20_17_58_+0100_=Q2bfUEBOTrwOu39 Content-Type: application/octet-stream; name="nfsclient.patch" Content-Disposition: attachment; filename="nfsclient.patch" Content-Transfer-Encoding: base64 LS0tIC91c3Ivc3JjL3N5cy9uZnNjbGllbnQvbmZzX3Zmc29wcy5jLm9yaWcJU2F0IE5vdiAyMiAw MzoyMTo0OSAyMDAzCisrKyAvdXNyL3NyYy9zeXMvbmZzY2xpZW50L25mc192ZnNvcHMuYwlUdWUg SmFuIDI3IDE5OjI2OjM1IDIwMDQKQEAgLTM1OSw2ICszNTksNyBAQAogCQltYXhmc2l6ZSA9IGZ4 ZHJfaHlwZXIoJmZzcC0+ZnNfbWF4ZmlsZXNpemUpOwogCQlpZiAobWF4ZnNpemUgPiAwICYmIG1h eGZzaXplIDwgbm1wLT5ubV9tYXhmaWxlc2l6ZSkKIAkJCW5tcC0+bm1fbWF4ZmlsZXNpemUgPSBt YXhmc2l6ZTsKKwkJbm1wLT5ubV9tb3VudHAtPm1udF9zdGF0LmZfaW9zaXplID0gbmZzX2lvc2l6 ZShubXApOwogCQlubXAtPm5tX3N0YXRlIHw9IE5GU1NUQV9HT1RGU0lORk87CiAJfQogCW1fZnJl ZW0obXJlcCk7CkBAIC03ODUsOCArNzg2LDEyIEBACiAKIAlubXAtPm5tX3RpbWVvID0gTkZTX1RJ TUVPOwogCW5tcC0+bm1fcmV0cnkgPSBORlNfUkVUUkFOUzsKLQlubXAtPm5tX3dzaXplID0gTkZT X1dTSVpFOwotCW5tcC0+bm1fcnNpemUgPSBORlNfUlNJWkU7CisJaWYgKChhcmdwLT5mbGFncyAm IE5GU01OVF9ORlNWMykgJiYgYXJncC0+c290eXBlID09IFNPQ0tfU1RSRUFNKSB7CisJCW5tcC0+ bm1fd3NpemUgPSBubXAtPm5tX3JzaXplID0gTkZTX01BWERBVEE7CisJfSBlbHNlIHsKKwkJbm1w LT5ubV93c2l6ZSA9IE5GU19XU0laRTsKKwkJbm1wLT5ubV9yc2l6ZSA9IE5GU19SU0laRTsKKwl9 CiAJbm1wLT5ubV9yZWFkZGlyc2l6ZSA9IE5GU19SRUFERElSU0laRTsKIAlubXAtPm5tX251bWdy cHMgPSBORlNfTUFYR1JQUzsKIAlubXAtPm5tX3JlYWRhaGVhZCA9IE5GU19ERUZSQUhFQUQ7CkBA IC04MzIsMTAgKzgzNywxNCBAQAogCSp2cHAgPSBORlNUT1YobnApOwogCiAJLyoKLQkgKiBHZXQg ZmlsZSBhdHRyaWJ1dGVzIGZvciB0aGUgbW91bnRwb2ludC4gIFRoaXMgaGFzIHRoZSBzaWRlCi0J ICogZWZmZWN0IG9mIGZpbGxpbmcgaW4gKCp2cHApLT52X3R5cGUgd2l0aCB0aGUgY29ycmVjdCB2 YWx1ZS4KKwkgKiBHZXQgZmlsZSBhdHRyaWJ1dGVzIGFuZCB0cmFuc2ZlciBwYXJhbWV0ZXJzIGZv ciB0aGUKKwkgKiBtb3VudHBvaW50LiAgVGhpcyBoYXMgdGhlIHNpZGUgZWZmZWN0IG9mIGZpbGxp bmcgaW4KKwkgKiAoKnZwcCktPnZfdHlwZSB3aXRoIHRoZSBjb3JyZWN0IHZhbHVlLgogCSAqLwot CVZPUF9HRVRBVFRSKCp2cHAsICZhdHRycywgY3VydGhyZWFkLT50ZF91Y3JlZCwgY3VydGhyZWFk KTsKKwlpZiAoYXJncC0+ZmxhZ3MgJiBORlNNTlRfTkZTVjMpCisJCW5mc19mc2luZm8obm1wLCAq dnBwLCBjdXJ0aHJlYWQtPnRkX3VjcmVkLCBjdXJ0aHJlYWQpOworCWVsc2UKKwkJVk9QX0dFVEFU VFIoKnZwcCwgJmF0dHJzLCBjdXJ0aHJlYWQtPnRkX3VjcmVkLCBjdXJ0aHJlYWQpOwogCiAJLyoK IAkgKiBMb3NlIHRoZSBsb2NrIGJ1dCBrZWVwIHRoZSByZWYuCkBAIC05MDUsNiArOTE0LDEzIEBA CiAJaWYgKGVycm9yKQogCQlyZXR1cm4gKGVycm9yKTsKIAl2cCA9IE5GU1RPVihucCk7CisJLyoK KwkgKiBHZXQgdHJhbnNmZXIgcGFyYW1ldGVycyBhbmQgYXR0cmlidXRlcyBmb3Igcm9vdCB2bm9k ZSBvbmNlLgorCSAqLworCWlmICgobm1wLT5ubV9zdGF0ZSAmIE5GU1NUQV9HT1RGU0lORk8pID09 IDAgJiYKKwkgICAgKG5tcC0+bm1fZmxhZyAmIE5GU01OVF9ORlNWMykpIHsKKwkJbmZzX2ZzaW5m byhubXAsIHZwLCBjdXJ0aHJlYWQtPnRkX3VjcmVkLCBjdXJ0aHJlYWQpOworCX0KIAlpZiAodnAt PnZfdHlwZSA9PSBWTk9OKQogCSAgICB2cC0+dl90eXBlID0gVkRJUjsKIAl2cC0+dl92ZmxhZyB8 PSBWVl9ST09UOwo= --Multipart=_Tue__27_Jan_2004_20_17_58_+0100_=Q2bfUEBOTrwOu39-- From owner-freebsd-performance@FreeBSD.ORG Tue Jan 27 13:07:07 2004 Return-Path: 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 39DEF16A4CF; Tue, 27 Jan 2004 13:07:07 -0800 (PST) Received: from obsecurity.dyndns.org (adsl-67-119-53-122.dsl.lsan03.pacbell.net [67.119.53.122]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9337F43D66; Tue, 27 Jan 2004 13:06:20 -0800 (PST) (envelope-from kris@obsecurity.org) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id D3B4B66C78; Tue, 27 Jan 2004 13:04:45 -0800 (PST) Date: Tue, 27 Jan 2004 13:04:45 -0800 From: Kris Kennaway To: Bj?rn Gr?nvall Message-ID: <20040127210445.GA68107@xor.obsecurity.org> References: <003c01c3de8d$d569edb0$471b3dd4@dual> <20040127201758.13be4f08.bg@sics.se> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="X1bOJ3K7DJ5YkBrT" Content-Disposition: inline In-Reply-To: <20040127201758.13be4f08.bg@sics.se> User-Agent: Mutt/1.4.1i X-Mailman-Approved-At: Tue, 27 Jan 2004 14:30:09 -0800 cc: performance@freebsd.org cc: Robert Watson cc: current@freebsd.org Subject: Re: Old SUN NFS performance papers. X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 27 Jan 2004 21:07:07 -0000 --X1bOJ3K7DJ5YkBrT Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, Jan 27, 2004 at 08:17:58PM +0100, Bj?rn Gr?nvall wrote: > On Sat, 24 Jan 2004 21:14:51 -0500 (EST) > Robert Watson wrote: >=20 > > I haven't done much benchmarking on NFS lately, but something worth > > remembering is that people have spent a lot of time researching and > > optimizing TCP for a variety of connection types, whereas the NFS code > > basically has a static implementation of RPC backoff and flow control t= hat > > hasn't evolved much. >=20 > One reason that FreeBSD users experience poor NFSv3/TCP performance is > that the defaults for rsize and wsize is unusually small, only > 8k. Solaris and HP-UX defaults to 32k for a good reason. I guess TCP > simply needs a little bit more data to chew on to be efficient. >=20 > I tested this on 5.2-CURRENT and found that large file read > performance went up from 56Mbit/s to 80Mbit/s, an improvement by 43%. >=20 > I have written a patch that makes FreeBSD use the same defaults as > Solaris and HP-UX. Note that with NFSv3 there is no risk associated > with specifying to large values for [rw]size. The server automatically > limits these values in the fsinfo rpc. Patch is attached. If no-one picks this up in the next few days, can you please send-pr it so it does not get lost? Kris --X1bOJ3K7DJ5YkBrT Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFAFtJtWry0BWjoQKURAkpxAJ43irNJqBozM1E1KwZuoBWxrTVf+gCgkxJA f4nz1j1BYHf53bMw0LzyHCE= =YdgB -----END PGP SIGNATURE----- --X1bOJ3K7DJ5YkBrT-- From owner-freebsd-performance@FreeBSD.ORG Wed Jan 28 02:26:52 2004 Return-Path: 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 99D8516A4CE; Wed, 28 Jan 2004 02:26:52 -0800 (PST) Received: from manian.sics.se (manian.sics.se [193.10.66.13]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9C6A743D62; Wed, 28 Jan 2004 02:26:38 -0800 (PST) (envelope-from bg@sics.se) Received: from sics.se (localhost [127.0.0.1]) by manian.sics.se (8.12.10/8.12.10) with SMTP id i0SAQWBS023884; Wed, 28 Jan 2004 11:26:32 +0100 (CET) (envelope-from bg@sics.se) Date: Wed, 28 Jan 2004 11:26:32 +0100 From: =?ISO-8859-1?Q?Bj=F6rn_Gr=F6nvall?= To: Kris Kennaway Message-Id: <20040128112632.6d040860.bg@sics.se> In-Reply-To: <20040127210445.GA68107@xor.obsecurity.org> References: <003c01c3de8d$d569edb0$471b3dd4@dual> <20040127201758.13be4f08.bg@sics.se> <20040127210445.GA68107@xor.obsecurity.org> Organization: SICS.SE X-Mailer: Sylpheed version 0.9.6claws (GTK+ 1.2.10; i386-portbld-freebsd5.2) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit cc: Robert Watson cc: bg@sics.se cc: performance@freebsd.org cc: current@freebsd.org Subject: Re: Old SUN NFS performance papers. X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 28 Jan 2004 10:26:52 -0000 On Tue, 27 Jan 2004 13:04:45 -0800 Kris Kennaway wrote: > If no-one picks this up in the next few days, can you please send-pr > it so it does not get lost? Done, kern/62024 (and an older kern/26324). Cheers, Bj鰎n -- _ _ ,_______________. Bjorn Gronvall (Bj鰎n Gr鰊vall) /_______________/| Swedish Institute of Computer Science | || PO Box 1263, S-164 29 Kista, Sweden | Schroedingers || Email: bg@sics.se, Phone +46 -8 633 15 25 | Cat |/ Cellular +46 -70 768 06 35, Fax +46 -8 751 72 30 '---------------' From owner-freebsd-performance@FreeBSD.ORG Sun Feb 1 16:46:56 2004 Return-Path: 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 9832716A4CE for ; Sun, 1 Feb 2004 16:46:56 -0800 (PST) Received: from f9.mail.ru (f9.mail.ru [194.67.57.39]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5EE5B43D39 for ; Sun, 1 Feb 2004 16:46:55 -0800 (PST) (envelope-from shmukler@mail.ru) Received: from mail by f9.mail.ru with local id 1AnSEo-000J3f-00 for freebsd-performance@freebsd.org; Mon, 02 Feb 2004 03:46:54 +0300 Received: from [24.184.137.78] by msg.mail.ru with HTTP; Mon, 02 Feb 2004 03:46:54 +0300 From: =?koi8-r?Q?=22?=Igor Shmukler=?koi8-r?Q?=22=20?= To: freebsd-performance@freebsd.org Mime-Version: 1.0 X-Mailer: mPOP Web-Mail 2.19 X-Originating-IP: [24.184.137.78] Date: Mon, 02 Feb 2004 03:46:54 +0300 Content-Type: text/plain; charset=koi8-r Content-Transfer-Encoding: 8bit Message-Id: Subject: realistic web benchmark X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: =?koi8-r?Q?=22?=Igor Shmukler=?koi8-r?Q?=22=20?= List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Feb 2004 00:46:56 -0000 I hope this is not entirely off topic. I am looking for some kind of realistic httpd performance benchmark. There are number of papers discussing how such benchmark could be constructed, but I was cusrious whether there is one available in a source form. IS. From owner-freebsd-performance@FreeBSD.ORG Sun Feb 1 16:57:48 2004 Return-Path: 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 126F616A4CE for ; Sun, 1 Feb 2004 16:57:48 -0800 (PST) Received: from hotmail.com (bay9-f31.bay9.hotmail.com [64.4.47.31]) by mx1.FreeBSD.org (Postfix) with ESMTP id D8AAA43D55 for ; Sun, 1 Feb 2004 16:57:39 -0800 (PST) (envelope-from aiwha@hotmail.com) Received: from mail pickup service by hotmail.com with Microsoft SMTPSVC; Sun, 1 Feb 2004 16:57:39 -0800 Received: from 218.2.157.46 by by9fd.bay9.hotmail.msn.com with HTTP; Mon, 02 Feb 2004 00:57:39 GMT X-Originating-IP: [218.2.157.46] X-Originating-Email: [aiwha@hotmail.com] X-Sender: aiwha@hotmail.com From: =?gb2312?B?us4gt7q0qA==?= To: freebsd-performance@freebsd.org Date: Mon, 02 Feb 2004 08:57:39 +0800 Mime-Version: 1.0 Content-Type: text/plain; charset=gb2312; format=flowed Message-ID: X-OriginalArrivalTime: 02 Feb 2004 00:57:39.0840 (UTC) FILETIME=[8E63B000:01C3E927] Subject: RE: realistic web benchmark X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Feb 2004 00:57:48 -0000 >From: "Igor Shmukler" >Reply-To: "Igor Shmukler" >To: freebsd-performance@freebsd.org >Subject: realistic web benchmark >Date: Mon, 02 Feb 2004 03:46:54 +0300 > >I hope this is not entirely off topic. > >I am looking for some kind of realistic httpd performance benchmark. There are >number of papers discussing how such benchmark could be constructed, but I was >cusrious whether there is one available in a source form. > >IS. ab is available in apache source code. ab stands for apache benchmark. _________________________________________________________________ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn From owner-freebsd-performance@FreeBSD.ORG Sun Feb 1 19:45:01 2004 Return-Path: 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 0755F16A4CE for ; Sun, 1 Feb 2004 19:45:01 -0800 (PST) Received: from fledge.watson.org (fledge.watson.org [204.156.12.50]) by mx1.FreeBSD.org (Postfix) with ESMTP id CFB3143D2D for ; Sun, 1 Feb 2004 19:44:59 -0800 (PST) (envelope-from robert@fledge.watson.org) Received: from fledge.watson.org (localhost [127.0.0.1]) by fledge.watson.org (8.12.10/8.12.10) with ESMTP id i123gTUd045650; Sun, 1 Feb 2004 22:42:29 -0500 (EST) (envelope-from robert@fledge.watson.org) Received: from localhost (robert@localhost)i123gSlk045647; Sun, 1 Feb 2004 22:42:29 -0500 (EST) (envelope-from robert@fledge.watson.org) Date: Sun, 1 Feb 2004 22:42:28 -0500 (EST) From: Robert Watson X-Sender: robert@fledge.watson.org To: =?koi8-r?Q?=22?=Igor Shmukler=?koi8-r?Q?=22=20?= In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: freebsd-performance@freebsd.org Subject: Re: realistic web benchmark X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Feb 2004 03:45:01 -0000 On Mon, 2 Feb 2004, [koi8-r] "Igor Shmukler[koi8-r] " wrote: > I hope this is not entirely off topic. > > I am looking for some kind of realistic httpd performance benchmark. > There are number of papers discussing how such benchmark could be > constructed, but I was cusrious whether there is one available in a > source form. In the past, I've used webstone for some web performance benchmarking. Recently, we've also been contacted with regard to a test suite named Web Polygraph. Robert N M Watson FreeBSD Core Team, TrustedBSD Projects robert@fledge.watson.org Senior Research Scientist, McAfee Research From owner-freebsd-performance@FreeBSD.ORG Sun Feb 1 20:23:22 2004 Return-Path: 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 544AD16A4CE; Sun, 1 Feb 2004 20:23:22 -0800 (PST) Received: from f21.mail.ru (f21.mail.ru [194.67.57.54]) by mx1.FreeBSD.org (Postfix) with ESMTP id 09F7443D41; Sun, 1 Feb 2004 20:23:21 -0800 (PST) (envelope-from shmukler@mail.ru) Received: from mail by f21.mail.ru with local id 1AnVcF-00095l-00; Mon, 02 Feb 2004 07:23:19 +0300 Received: from [24.184.137.78] by msg.mail.ru with HTTP; Mon, 02 Feb 2004 07:23:19 +0300 From: =?koi8-r?Q?=22?=Igor Shmukler=?koi8-r?Q?=22=20?= To: =?koi8-r?Q?=22?=Robert Watson=?koi8-r?Q?=22=20?= Mime-Version: 1.0 X-Mailer: mPOP Web-Mail 2.19 X-Originating-IP: [24.184.137.78] Date: Mon, 02 Feb 2004 07:23:19 +0300 In-Reply-To: Content-Type: text/plain; charset=koi8-r Content-Transfer-Encoding: 8bit Message-Id: cc: freebsd-performance@freebsd.org Subject: Re[2]: realistic web benchmark X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: =?koi8-r?Q?=22?=Igor Shmukler=?koi8-r?Q?=22=20?= List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Feb 2004 04:23:22 -0000 > > I hope this is not entirely off topic. > > > > I am looking for some kind of realistic httpd performance benchmark. > > There are number of papers discussing how such benchmark could be > > constructed, but I was cusrious whether there is one available in a > > source form. > > In the past, I've used webstone for some web performance benchmarking. > Recently, we've also been contacted with regard to a test suite named Web > Polygraph. I probably was not 100% clear when I asked the question. Sorry. I am looking for test that would tell me how will server perform in the field on WAN. Webstone is a bit naive about benchmarking. It allows comparison of server scores, however the fact that a particular OS/server combination scores higher does not guaranatee that it will have higher thoughput in real life situation. I scanned briefly through Web Polygraph's documentation and it seems like more powerful tool, but I could not find how to emulate large delays and packet loss common for WANs. IS. From owner-freebsd-performance@FreeBSD.ORG Mon Feb 2 08:56:22 2004 Return-Path: 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 C77E016A4E6 for ; Mon, 2 Feb 2004 08:56:22 -0800 (PST) Received: from Princeton.EDU (postoffice01.Princeton.EDU [128.112.129.75]) by mx1.FreeBSD.org (Postfix) with ESMTP id 606AB43D46 for ; Mon, 2 Feb 2004 08:56:20 -0800 (PST) (envelope-from yruan@cs.princeton.edu) Received: from smtpserver1.Princeton.EDU (smtpserver1.Princeton.EDU [128.112.129.65]) by Princeton.EDU (8.12.9/8.12.9) with ESMTP id i12GuJVg029211; Mon, 2 Feb 2004 11:56:19 -0500 (EST) Received: from cs.princeton.edu (targe.CS.Princeton.EDU [128.112.139.194]) (authenticated bits=0)i12GuFSu006136 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NOT); Mon, 2 Feb 2004 11:56:18 -0500 (EST) Message-ID: <401E8106.674C9EDB@cs.princeton.edu> Date: Mon, 02 Feb 2004 11:55:34 -0500 From: Yaoping Ruan X-Mailer: Mozilla 4.76 [en] (Windows NT 5.0; U) X-Accept-Language: en MIME-Version: 1.0 To: Igor Shmukler References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit cc: freebsd-performance@freebsd.org Subject: Re: realistic web benchmark X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 02 Feb 2004 16:56:23 -0000 "Igor Shmukler" wrote: > I probably was not 100% clear when I asked the question. Sorry. I am looking for > test that would tell me how will server perform in the field on WAN. > Erich Nahum at IBM Research had a work called WASP. They used Dummynet as WAN environment emulator, and s-client as WWW request generator. The paper is at: http://www.research.ibm.com/people/n/nahum/publications/sigmetrics01-wasp.pdf S-client is available at: http://www.cs.rice.edu/CS/Systems/Web-measurement/sources.html Hope this is helpful. - Yaoping From owner-freebsd-performance@FreeBSD.ORG Tue Feb 3 02:45:30 2004 Return-Path: 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 69C8A16A4CE; Tue, 3 Feb 2004 02:45:30 -0800 (PST) Received: from stork.mail.pas.earthlink.net (stork.mail.pas.earthlink.net [207.217.120.188]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1574543D46; Tue, 3 Feb 2004 02:45:28 -0800 (PST) (envelope-from tlambert2@mindspring.com) Received: from user-2ivfmse.dialup.mindspring.com ([165.247.219.142] helo=mindspring.com) by stork.mail.pas.earthlink.net with asmtp (SSLv3:RC4-MD5:128) (Exim 3.33 #1) id 1Any3W-0000IS-00; Tue, 03 Feb 2004 02:45:23 -0800 Message-ID: <401F7BEA.566F9342@mindspring.com> Date: Tue, 03 Feb 2004 02:46:02 -0800 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: Igor Shmukler References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-ELNK-Trace: b1a02af9316fbb217a47c185c03b154d40683398e744b8a42b74c05c75a0af5847ee5f5f9327823c350badd9bab72f9c350badd9bab72f9c350badd9bab72f9c cc: freebsd-performance@freebsd.org cc: Robert Watson Subject: Re: realistic web benchmark X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Feb 2004 10:45:30 -0000 "Igor Shmukler" wrote: > > In the past, I've used webstone for some web performance benchmarking. > > Recently, we've also been contacted with regard to a test suite named Web > > Polygraph. > > I probably was not 100% clear when I asked the question. Sorry. I am looking for > test that would tell me how will server perform in the field on WAN. > > Webstone is a bit naive about benchmarking. It allows comparison of server > scores, however the fact that a particular OS/server combination scores higher > does not guaranatee that it will have higher thoughput in real life situation. > > I scanned briefly through Web Polygraph's documentation and it seems like more > powerful tool, but I could not find how to emulate large delays and packet loss > common for WANs. That's what "dummynet" does for you: it lets you simulate lossy traffic, etc.; I presonally don't find it to be that meaningful, since the only thing that's actually going to see lossy traffic is the L4/L7 load balancer sitting in front of your server farm, so unless you are testing your load balancer, you really aren't going to get much information out of a lossy network test. Also, your most critical point of failure is going to be your really slow client connections taking up all your sockets and thus starving your faster client connections by being in the way. If you care about comparison points, webstone and Microsoft's WAST tool are going to be what people will be comparing you against when they run their own tests on an evaluation. http://www.mindcraft.com/webstone/ http://www.microsoft.com/technet/itsolutions/intranet/downloads/webstres.asp The WAST application from Microsoft has the negative effect that it tends to fill up your server with packets in FIN-WAIT-2 state. This is because it doesn't do the full handshake on shutdown of connections, it just RST's them. The problem with this approach is that if you lose an RST packet, as opposed to a FIN, you aren't going to get a retransmit in 2 MSL. Julian Elisher did some patches for the TCP stack while at Whistle that address this issue the right way, by pretending to not have got the FIN that sent you from FIN-WAIT-1 to FIN-WAIT-2. This is also the same fix that Windows NT uses. If you don't make this fix, then you should expect that your server will "fill up" with idle connections when you are running WAST, and eventually stop serving pages. The primary value in Polygraph is stress-testing. It's mostly for proxy servers, and its main value lies in it setting up cache-busting scenarios by pre-loading a "hot" cache to force the limitation to be your proxy server, load balancer, or whatever. Mostly as a statement about the undesirability of such devices inre: the end-to-end nature of the net "as it's supposed to be". http://www.web-polygraph.org/ It's actually really amusing that Network Appliance "cheats" on the Polygraph benchmarks with their caching proxy appliance by doing random page replacement in order to defeat the Polygraph attempt to defeat the proxy cache by guessing it's size, and then making the workload such that there's 100% cache misses. That you could have such a pessimal algorithm, and that that was the best way to get a good score on the benchmark, to me, says a lot about the actual value of the benchmark as a benchmark. If you are just into stress testing, rather than running Polygraph, you'd likely be better off running http_load on a bunch of boxes: http://www.acme.com/software/http_load/ Or if you aren't adverse to a commercial box, you might want to consider the Web Avalanche products from Spirent, which they got in their acquisition of Caw Networks: http://www.caw.com/ It's basically a single box that's the equivalent of a lab worth of UNIX systems running a bunch of copies of http_load. It's very good at making web servers and proxies, etc., fall over dead if they have any issues at all. -- Terry From owner-freebsd-performance@FreeBSD.ORG Tue Feb 3 09:29:33 2004 Return-Path: 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 44F2C16A4CE for ; Tue, 3 Feb 2004 09:29:33 -0800 (PST) Received: from mail.ubergeeks.com (lorax.ubergeeks.com [209.145.65.55]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6519243D2D for ; Tue, 3 Feb 2004 09:29:31 -0800 (PST) (envelope-from adrian+freebsd-perf@ubergeeks.com) Received: from mail.ubergeeks.com (localhost [127.0.0.1]) by mail.ubergeeks.com (8.12.9p2/8.12.9) with ESMTP id i13HQJqZ025727 for ; Tue, 3 Feb 2004 12:26:19 -0500 (EST) (envelope-from adrian+freebsd-perf@ubergeeks.com) Received: from localhost (adrian@localhost)i13HQJxC025724 for ; Tue, 3 Feb 2004 12:26:19 -0500 (EST) (envelope-from adrian+freebsd-perf@ubergeeks.com) X-Authentication-Warning: lorax.ubergeeks.com: adrian owned process doing -bs Date: Tue, 3 Feb 2004 12:26:19 -0500 (EST) From: Adrian Filipi Sender: adrian@ubergeeks.com To: freebsd-performance@freebsd.org Message-ID: <20040201184719.S3066@lorax.ubergeeks.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-MailScanner-Information: Please contact the ISP for more information X-MailScanner: Found to be clean Subject: frustratingly slow box at 4GB, but not 1GB of memory X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Feb 2004 17:29:33 -0000 Hi folks, I've got a box I'm helping tune and it's not really responding as I'd expect. It's a dual 2GHz Xeon box running on a Supermicro P4DLR motherboard with 4GB of physical RAM. It has 4GB of swap configured on a hardware RAID5 configuration. It's running 4.9-RELEASE. It's also running Apache with PHP and Perl CGI's. The CGI's read/write to a MySQL DB. Briefly, it responds pretty quickly when I only let the kernel use 1GB of the RAM, but gets bogged down when I let it use the full 4GB. I'm limiting the RAM by setting hw.physmem. I've tried some of the things that I could dig up in the archives, but the knowledge is pretty widely scattered. :-( I have MAXUSERS set to 96 as that gets me plenty of file descriptors and pid's. While tight, I believe I could get away with only 32, as there should not be more than 300-400 processes running at any one time. I set vm.pmap.shpgperproc="300" because of warnings from the kernel. I'm not sure if this is the best value, but it made the warnings go away. First, upping the KVA_PAGES to 768 to achieve a 3:1 ration of kernel to user memory, didn't change things. Disabling swap seems to make a huge difference. It changes the time results for "time top -bu" by two orders of magnitude from 0.006s to 0.6s! In fact, I can run the box reasonably well with 2GB of RAM with swap disabled. If I let the box boot using all of the RAM, it is very slow. e.g. "time top -bu >/dev/null" takes about 6 seconds. I am seeing large runs of time (15-30 seconds) where the system time usage is in the high range (75-99). Does anyone have a suggestion on how to determine which kernel data structures are the problems? thanks, Adrian -- [ adrian@ubergeeks.com ] From owner-freebsd-performance@FreeBSD.ORG Fri Feb 6 09:38:01 2004 Return-Path: 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 D2FB416A4CE for ; Fri, 6 Feb 2004 09:38:01 -0800 (PST) Received: from mail2.dbitech.ca (radius.wavefire.com [64.141.13.252]) by mx1.FreeBSD.org (Postfix) with SMTP id 9CC9343D75 for ; Fri, 6 Feb 2004 09:37:44 -0800 (PST) (envelope-from darcy@wavefire.com) Received: (qmail 13081 invoked from network); 6 Feb 2004 18:23:28 -0000 Received: from dbitech.wavefire.com (HELO 64.141.15.253) (darcy@64.141.15.253) by radius.wavefire.com with SMTP; 6 Feb 2004 18:23:28 -0000 From: Darcy Buskermolen Organization: Wavefire Technologies Corp. To: Soeren Schmidt , Trent George Date: Fri, 6 Feb 2004 09:37:22 -0800 User-Agent: KMail/1.5.4 References: <200307180815.h6I8FnLd071918@spider.deepcore.dk> In-Reply-To: <200307180815.h6I8FnLd071918@spider.deepcore.dk> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Message-Id: <200402060937.22213.darcy@wavefire.com> cc: freebsd-performance@freebsd.org Subject: Re: promise sx4000 card & tx4000 results over 170mb/s sustained read X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 06 Feb 2004 17:38:01 -0000 On July 18, 2003 01:15 am, Soeren Schmidt wrote: > > Question #2 > > > > I purchased a sx4000 card in the hope to increase speed :-) > > I noticed that PDC20621 is not in ata-chipset.c yet > > The card is not recognised on boot ether. > > > > Is there anything I can do or provide to help you add support for this > > card ? > > I do have the sx4000 in the works (thanks to Promise who has provided docs > and HW to make this happen).. > > -S=F8ren I also have an sx4000 card, and I'm wondering how support for this item is= =20 comming along. if you requre any help in coding or testing, please feel fre= e=20 to let me know as I'd like to deploy this card natily under FreeBSD instead= =20 of over NFS from a linux box where it is now. =2D-=20 Darcy Buskermolen Wavefire Technologies Corp. ph: 250.717.0200 fx: 250.763.1759 http://www.wavefire.com From owner-freebsd-performance@FreeBSD.ORG Mon Feb 2 20:08:51 2004 Return-Path: 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 B0F4B16A4CE for ; Mon, 2 Feb 2004 20:08:51 -0800 (PST) Received: from web60806.mail.yahoo.com (web60806.mail.yahoo.com [216.155.196.69]) by mx1.FreeBSD.org (Postfix) with SMTP id A842C43D3F for ; Mon, 2 Feb 2004 20:08:50 -0800 (PST) (envelope-from richard_bejtlich@yahoo.com) Message-ID: <20040203040849.57265.qmail@web60806.mail.yahoo.com> Received: from [68.84.6.72] by web60806.mail.yahoo.com via HTTP; Mon, 02 Feb 2004 20:08:49 PST Date: Mon, 2 Feb 2004 20:08:49 -0800 (PST) From: Richard Bejtlich To: freebsd-performance@freebsd.org MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailman-Approved-At: Fri, 06 Feb 2004 12:47:03 -0800 Subject: OSU Open Source Lab Benchmark Challenge X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 03 Feb 2004 04:08:51 -0000 Hello, Did anyone else read about the OSU Open Source Lab Benchmark Challenge (http://osuosl.org/benchmarks/bc/) at Slashdot today? http://developers.slashdot.org/article.pl?sid=04/02/03/0021232&mode=nested&tid=106&tid=117&tid=185&tid=190&tid=99 It appears to be a good-faith attempt to compare the performance of various open source operating systems. This could be a chance for the world to see FreeBSD shine. Does anyone plan to help "tweak" the FreeBSD install, as allowed by the contest? I'm interested in this as a FreeBSD user and advocate, but admit not being sufficiently advanced to make performance recommendations. Sincerely, Richard http://www.taosecurity.com __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free web site building tool. Try it! http://webhosting.yahoo.com/ps/sb/ From owner-freebsd-performance@FreeBSD.ORG Fri Feb 6 20:02:17 2004 Return-Path: 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 65E2D16A4CE for ; Fri, 6 Feb 2004 20:02:17 -0800 (PST) Received: from mta7.pltn13.pbi.net (mta7.pltn13.pbi.net [64.164.98.8]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3180543D1F for ; Fri, 6 Feb 2004 20:02:16 -0800 (PST) (envelope-from todd@qdsdirect.com) Received: from qdsdirect.com (adsl-65-68-200-81.dsl.rcsntx.swbell.net [65.68.200.81])id i1742B9D005991 for ; Fri, 6 Feb 2004 20:02:15 -0800 (PST) Message-ID: <4025B3BE.3020009@qdsdirect.com> Date: Sat, 07 Feb 2004 21:57:50 -0600 From: Todd Lewis User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.6b) Gecko/20031205 Thunderbird/0.4 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 Subject: Raid 5 performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 07 Feb 2004 04:02:17 -0000 I am using FreeBSD 4.9, with a 3ware RAID 5 1 gig memory 2.8g p4 Three questions. 1. FreeBSD has a 16k block size. The RAID card is set at 64k Block size(its sweet spot). My logic tells me that increasing the block size to 64k would increase disk read and write access. But, everything I read suggest going above 64k is dangerous. Are their any recomendations on performance a stability concerns when increasnig the block size to 64k when using a RAID controler. 2. The vfs.hirunningspace variable defualts to 1meg. From what I've read this looks like a buffer. I'm guessing that its set to 1meg becuase most drives have 1~2 megs of memory. So following that logic and with safety in mind. For drives with 4 megs cache, I would set vfs.hirunningspace to 2 megs. 8megs of cache 4 megs to vfs.hirunningspace. So, my 64 megs raid control would have a vfs.hirunningspace 32. 3. Any other stellar performance info would be appriciated. netstat -m mbuf usage: GEN list: 4/480 (in use/in pool) CPU #0 list: 443/656 (in use/in pool) Total: 447/1136 (in use/in pool) Maximum number allowed on each CPU list: 512 Maximum possible: 51200 Allocated mbuf types: 358 mbufs allocated to data 89 mbufs allocated to packet headers 2% of mbuf map consumed mbuf cluster usage: GEN list: 12/564 (in use/in pool) CPU #0 list: 263/304 (in use/in pool) Total: 275/868 (in use/in pool) Maximum number allowed on each CPU list: 128 Maximum possible: 25600 3% of cluster map consumed 2020 KBytes of wired memory reserved (32% in use) 0 requests for memory denied 0 requests for memory delayed 0 calls to protocol drain routines dd if=/dev/da0s1f of=/dev/null bs=1m count=100 code: 100+0 records in 100+0 records out 104857600 bytes transferred in 6.331798 secs (16560478 bytes/sec) From owner-freebsd-performance@FreeBSD.ORG Sat Feb 7 01:39:41 2004 Return-Path: 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 2F15716A4CE for ; Sat, 7 Feb 2004 01:39:41 -0800 (PST) Received: from geminix.org (gen129.n001.c02.escapebox.net [213.73.91.129]) by mx1.FreeBSD.org (Postfix) with ESMTP id E568843D1D for ; Sat, 7 Feb 2004 01:39:40 -0800 (PST) (envelope-from gemini@geminix.org) Message-ID: <4024B259.8010005@geminix.org> Date: Sat, 07 Feb 2004 10:39:37 +0100 From: Uwe Doering Organization: Private UNIX Site User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.6) Gecko/20040119 X-Accept-Language: en-us, en MIME-Version: 1.0 To: freebsd-performance@freebsd.org References: <4025B3BE.3020009@qdsdirect.com> In-Reply-To: <4025B3BE.3020009@qdsdirect.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Received: from gemini by geminix.org with asmtp (TLSv1:AES256-SHA:256) (Exim 3.36 #1) id 1ApOw7-000Jua-00; Sat, 07 Feb 2004 10:39:39 +0100 Subject: Re: Raid 5 performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 07 Feb 2004 09:39:41 -0000 Todd Lewis wrote: > I am using FreeBSD 4.9, with a 3ware RAID 5 > 1 gig memory 2.8g p4 > > Three questions. > > 1. FreeBSD has a 16k block size. The RAID card is set at 64k > Block size(its sweet spot). My logic tells me that > increasing the block size to 64k would increase disk > read and write access. But, everything I read suggest > going above 64k is dangerous. Are their any recomendations > on performance a stability concerns when increasnig the > block size to 64k when using a RAID controler. A RAID controller normally has nothing to do with the file system's block size. Are you sure that you're not mixing this up with the stripe size? Which stripe size to use with a RAID controller depends on your performance priorities. If there are a lot of concurrent disk operations a larger stripe size is better because then a single disk operation tends to be limited to only one disk drive, leaving the remaining drives free to perform other and possibly unrelated disk operations at the same time. On the other hand, if sequential i/o throughput is important a smaller stripe size is better. > 2. The vfs.hirunningspace variable defualts to 1meg. From what I've > read this looks like a buffer. I'm guessing that its set to > 1meg becuase most drives have 1~2 megs of memory. So following > that logic and with safety in mind. For drives with 4 megs > cache, I would set vfs.hirunningspace to 2 megs. 8megs of cache > 4 megs to vfs.hirunningspace. So, my 64 megs raid control would > have a vfs.hirunningspace 32. It is my experience, too, that this variable is too low by default for "intelligent" disk controllers with large buffers. However, the amount of buffer space for outstanding disk operations is taken from the kernel's disk i/o buffer, which is normally auto-sized at boot time, based on the amount of memory you have. But you can also override it. You may want to check 'vfs.maxbufspace' and make 'vfs.hirunningspace' only a fraction of it. Not more that 1/4, for instance. And adapting 'vfs.lorunningspace' accordingly is also a good idea (it's a hysteresis). Uwe -- Uwe Doering | EscapeBox - Managed On-Demand UNIX Servers gemini@geminix.org | http://www.escapebox.net From owner-freebsd-performance@FreeBSD.ORG Sat Feb 7 02:30:23 2004 Return-Path: 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 D0E7E16A4CF for ; Sat, 7 Feb 2004 02:30:23 -0800 (PST) Received: from mta8.srv.hcvlny.cv.net (mta8.srv.hcvlny.cv.net [167.206.5.75]) by mx1.FreeBSD.org (Postfix) with ESMTP id B9FD743D1D for ; Sat, 7 Feb 2004 02:30:23 -0800 (PST) (envelope-from all@biosys.net) Received: from megalomaniac.biosys.net (ool-43529b09.dyn.optonline.net [67.82.155.9]) by mta8.srv.hcvlny.cv.net (iPlanet Messaging Server 5.2 HotFix 1.16 (built May 14 2003)) with ESMTP id <0HSP00L9LMI0K0@mta8.srv.hcvlny.cv.net> for freebsd-performance@freebsd.org; Sat, 07 Feb 2004 05:30:00 -0500 (EST) Date: Sat, 07 Feb 2004 05:35:25 -0500 From: Allen Landsidel In-reply-to: <4024B259.8010005@geminix.org> X-Sender: bsdasym@pop.hotpop.com To: freebsd-performance@freebsd.org Message-id: <6.0.0.22.2.20040207051454.03517668@pop.hotpop.com> MIME-version: 1.0 X-Mailer: QUALCOMM Windows Eudora Version 6.0.0.22 Content-type: text/plain; charset=us-ascii; format=flowed Content-transfer-encoding: 7BIT References: <4025B3BE.3020009@qdsdirect.com> <4024B259.8010005@geminix.org> Subject: Re: Raid 5 performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 07 Feb 2004 10:30:24 -0000 At 04:39 2/7/2004, Uwe Doering wrote: >Todd Lewis wrote: >>I am using FreeBSD 4.9, with a 3ware RAID 5 >>1 gig memory 2.8g p4 >>Three questions. >>1. FreeBSD has a 16k block size. The RAID card is set at 64k >> Block size(its sweet spot). My logic tells me that >> increasing the block size to 64k would increase disk >> read and write access. But, everything I read suggest >> going above 64k is dangerous. Are their any recomendations >> on performance a stability concerns when increasnig the >> block size to 64k when using a RAID controler. > >A RAID controller normally has nothing to do with the file system's block >size. Are you sure that you're not mixing this up with the stripe >size? Which stripe size to use with a RAID controller depends on your >performance priorities. > >If there are a lot of concurrent disk operations a larger stripe size is >better because then a single disk operation tends to be limited to only >one disk drive, leaving the remaining drives free to perform other and >possibly unrelated disk operations at the same time. On the other hand, >if sequential i/o throughput is important a smaller stripe size is better. I am compelled to step up here and say that this flies in the face not only of everything I have read, ever, about RAID -- but my own personal experiences as well on a variety of controllers and drives ranging from ATA highpoint/maxtor combos up to 4ch u160 ICP-Vortex/15krpm monsters. My experience has told me that for mostly sequential I/O, bigger is better, up to a point. 128KB stripes are much faster than 16KB stripes on every combination I've ever used when it comes to sequential I/O. Random I/O tends to prefer smaller stripe sizes because as you said, spreading things out over disks is better, and in random I/O the request sizes tend to be much smaller. My advice to our intrepid advice seeker is simply this. Get your system running under a load pattern as close to 'normal' as you can get. Then, run "systat -vmstat 1" and watch the output. Down with the disks is a row "KB/t" which is Kilobytes per transaction. Keep an eye on this number, and note the approximate value that seems to occur most of the time. Then when building your array, set the stripe size to the smallest value you can that is this size or larger. This will ensure that the majority of your disk activity hits every disk, getting you maximum throughput. Set the value too low and every disk will be hit more than once for one request which means more controller overhead. Set it too high and you'll hit fewer (perhaps only one) disk on the majority of your requests, negating any performance benefit RAID could offer you. If you plan to expand the array later by adding disks, take that into account and use a slightly smaller stripe size -- disks will be hit more than once until you add more, at which point you'll get closer to each disk only getting one request per transaction, which is the sweet spot for non-sequential I/O which is what most production servers (especially databases) do, most of the time. I agree with one thing said here however -- the OS block size has little to nothing to do with this except for one thing -- you can be almost certain that your average request size will be equal to or some multiple of this value. From owner-freebsd-performance@FreeBSD.ORG Sat Feb 7 07:14:12 2004 Return-Path: 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 4EBC416A4CE for ; Sat, 7 Feb 2004 07:14:12 -0800 (PST) Received: from da.mailomat.net (bn.mailomat.net [212.63.50.11]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6A2EF43D1D for ; Sat, 7 Feb 2004 07:14:10 -0800 (PST) (envelope-from ap@bnc.net) Received: This line has been intentionally left blank. Received: from bnc.net (246kmldlynbcpt6z@port-212-202-15-224.reverse.qsc.de [212.202.15.224]) (user=bnc.mail mech=LOGIN bits=0) i17FDajg084120 for ; Sat, 7 Feb 2004 16:13:37 +0100 (CET) (envelope-from ap@bnc.net) Received: by bnc.net (CommuniGate Pro PIPE 4.1.8) with PIPE id 495377; Sat, 07 Feb 2004 16:13:34 +0100 Received: from [194.39.192.247] (account ap HELO [194.39.192.247]) by bnc.net (CommuniGate Pro SMTP 4.1.8) with ESMTP-TLS id 495379 for freebsd-performance@freebsd.org; Sat, 07 Feb 2004 16:12:59 +0100 Mime-Version: 1.0 (Apple Message framework v613) In-Reply-To: <4024B259.8010005@geminix.org> References: <4025B3BE.3020009@qdsdirect.com> <4024B259.8010005@geminix.org> X-Priority: 3 Content-Type: text/plain; charset=US-ASCII; format=flowed Message-Id: <1CDE346A-5980-11D8-94F9-000A95A0BB90@bnc.net> Content-Transfer-Encoding: 7bit From: Achim Patzner Date: Sat, 7 Feb 2004 16:12:58 +0100 To: freebsd-performance@freebsd.org X-Mailer: Apple Mail (2.613) X-MailScanner-Information: Please contact info@mailomat.net for more information X-MailScanner: Found to be clean X-Virus-Scanned: by AMaViS perl-11 Subject: Re: Raid 5 performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 07 Feb 2004 15:14:12 -0000 Am 07.02.2004 um 10:39 schrieb Uwe Doering: >> 1. FreeBSD has a 16k block size. The RAID card is set at 64k >> Block size(its sweet spot). > A RAID controller normally has nothing to do with the file system's > block size. Are you sure that you're not mixing this up with the > stripe size? He is. > Which stripe size to use with a RAID controller depends on your > performance priorities. I just (painfully) found out that at least for 3ware controllers the most important part of getting performance out of them is picking the right firmware... Right now we're back at "the more recent the better". Did anyone already try the 3ware SATA boards? Achim From owner-freebsd-performance@FreeBSD.ORG Sat Feb 7 10:34:24 2004 Return-Path: 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 5ADAE16A4CE for ; Sat, 7 Feb 2004 10:34:24 -0800 (PST) Received: from rms04.rommon.net (rms04.rommon.net [212.54.2.140]) by mx1.FreeBSD.org (Postfix) with ESMTP id AC89943D1D for ; Sat, 7 Feb 2004 10:34:23 -0800 (PST) (envelope-from pete@he.iki.fi) Received: from he.iki.fi (h81.vuokselantie10.fi [193.64.42.129]) by rms04.rommon.net (8.12.9p1/8.12.9) with ESMTP id i17IYJW7052209; Sat, 7 Feb 2004 20:34:19 +0200 (EET) (envelope-from pete@he.iki.fi) Message-ID: <40252F70.3010805@he.iki.fi> Date: Sat, 07 Feb 2004 20:33:20 +0200 From: Petri Helenius User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.6) Gecko/20040113 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Achim Patzner References: <4025B3BE.3020009@qdsdirect.com> <4024B259.8010005@geminix.org> <1CDE346A-5980-11D8-94F9-000A95A0BB90@bnc.net> In-Reply-To: <1CDE346A-5980-11D8-94F9-000A95A0BB90@bnc.net> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit cc: freebsd-performance@freebsd.org Subject: Re: Raid 5 performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 07 Feb 2004 18:34:24 -0000 Achim Patzner wrote: > > Did anyone already try the 3ware SATA boards? > The dmesg does not tell the difference but the 12 port SATA card works; twe0: <3ware 7000 series Storage Controller. Driver version 1.50.00.000> port 0x7000-0x700f mem 0xf0800000 -0xf0ffffff,0xf0200000-0xf020000f irq 28 at device 2.0 on pci3 twe0: 12 ports, Firmware FE7S 1.05.00.065, BIOS BE7X 1.08.00.048 Pete From owner-freebsd-performance@FreeBSD.ORG Sat Feb 7 15:36:56 2004 Return-Path: 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 EFADD16A4CE for ; Sat, 7 Feb 2004 15:36:55 -0800 (PST) Received: from geminix.org (gen129.n001.c02.escapebox.net [213.73.91.129]) by mx1.FreeBSD.org (Postfix) with ESMTP id 424D843D1F for ; Sat, 7 Feb 2004 15:36:55 -0800 (PST) (envelope-from gemini@geminix.org) Message-ID: <40257693.7080308@geminix.org> Date: Sun, 08 Feb 2004 00:36:51 +0100 From: Uwe Doering Organization: Private UNIX Site User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.6) Gecko/20040119 X-Accept-Language: en-us, en MIME-Version: 1.0 To: freebsd-performance@freebsd.org References: <4025B3BE.3020009@qdsdirect.com> <4024B259.8010005@geminix.org> <6.0.0.22.2.20040207051454.03517668@pop.hotpop.com> In-Reply-To: <6.0.0.22.2.20040207051454.03517668@pop.hotpop.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Received: from gemini by geminix.org with asmtp (TLSv1:AES256-SHA:256) (Exim 3.36 #1) id 1Apc0M-000Ams-00; Sun, 08 Feb 2004 00:36:54 +0100 Subject: Re: Raid 5 performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 07 Feb 2004 23:36:56 -0000 Allen Landsidel wrote: > At 04:39 2/7/2004, Uwe Doering wrote: >> Todd Lewis wrote: >>> I am using FreeBSD 4.9, with a 3ware RAID 5 >>> 1 gig memory 2.8g p4 >>> Three questions. >>> 1. FreeBSD has a 16k block size. The RAID card is set at 64k >>> Block size(its sweet spot). My logic tells me that >>> increasing the block size to 64k would increase disk >>> read and write access. But, everything I read suggest >>> going above 64k is dangerous. Are their any recomendations >>> on performance a stability concerns when increasnig the >>> block size to 64k when using a RAID controler. >> >> A RAID controller normally has nothing to do with the file system's >> block size. Are you sure that you're not mixing this up with the >> stripe size? Which stripe size to use with a RAID controller depends >> on your performance priorities. >> >> If there are a lot of concurrent disk operations a larger stripe size >> is better because then a single disk operation tends to be limited to >> only one disk drive, leaving the remaining drives free to perform >> other and possibly unrelated disk operations at the same time. On the >> other hand, if sequential i/o throughput is important a smaller stripe >> size is better. > > I am compelled to step up here and say that this flies in the face not > only of everything I have read, ever, about RAID -- but my own personal > experiences as well on a variety of controllers and drives ranging from > ATA highpoint/maxtor combos up to 4ch u160 ICP-Vortex/15krpm monsters. > > My experience has told me that for mostly sequential I/O, bigger is > better, up to a point. 128KB stripes are much faster than 16KB stripes > on every combination I've ever used when it comes to sequential I/O. > > Random I/O tends to prefer smaller stripe sizes because as you said, > spreading things out over disks is better, and in random I/O the request > sizes tend to be much smaller. I didn't talk about random I/O on a single file but about parallel, unrelated disk operations (by different processes, for instance), each of which may very well be part of a sequential read or write. FreeBSD does clustering, so when an I/O operation actually takes place it can be much larger than the 16k block size. So in order to achieve a sufficiently high statistical likelyhood that an I/O operation is limited to a single disk, leaving the remaining disks free for other I/O, you need a stripe size considerably larger than the file system's block size. A file server with many parallel, unrelated accesses by NFS clients is a good example. There you certainly cannot generalize that files and therefore (clustered) I/O operations tend to be small. This generalization is only valid to some extent in case a file is a database where the I/O operations are mostly random and indeed tend to be small. What all this means is that you need to carefully analyse the intended use and access pattern in order to pick the right stripe size, since there is no one-size-fits-all. Uwe -- Uwe Doering | EscapeBox - Managed On-Demand UNIX Servers gemini@geminix.org | http://www.escapebox.net