From owner-freebsd-scsi Sun Feb 15 01:59:55 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id BAA04908 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 01:59:55 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from news1.gtn.com (news1.gtn.com [192.109.159.3]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id BAA04894 for ; Sun, 15 Feb 1998 01:59:47 -0800 (PST) (envelope-from andreas@klemm.gtn.com) Received: (from uucp@localhost) by news1.gtn.com (8.8.6/8.8.6) with UUCP id KAA00563; Sun, 15 Feb 1998 10:45:07 +0100 (MET) Received: (from andreas@localhost) by klemm.gtn.com (8.8.8/8.8.7) id KAA09746; Sun, 15 Feb 1998 10:28:05 +0100 (CET) (envelope-from andreas) Message-ID: <19980215102805.21437@klemm.gtn.com> Date: Sun, 15 Feb 1998 10:28:05 +0100 From: Andreas Klemm To: "Justin T. Gibbs" Cc: scsi@FreeBSD.ORG Subject: Re: Amanda 2.4b6 port References: <19980214180849.43287@klemm.gtn.com> <199802150706.AAA16260@narnia.plutotech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <199802150706.AAA16260@narnia.plutotech.com>; from Justin T. Gibbs on Sun, Feb 15, 1998 at 12:06:58AM -0700 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > Justin, please help ;-) > Run CAM. How available please ? -- Andreas Klemm powered by ,,symmetric multiprocessor FreeBSD'' To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 02:43:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id CAA12030 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 02:43:09 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from tapir.noc.fh-lippe.de (tapir.noc.fh-lippe.de [193.16.112.7]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id CAA12018; Sun, 15 Feb 1998 02:43:06 -0800 (PST) (envelope-from lkoeller@cc.FH-Lippe.DE) Received: from spock.cc.fh-lippe.de by tapir.noc.fh-lippe.de with smtp (Smail3.1.28.1 #25) id m0y41XH-0001QGC; Sun, 15 Feb 98 11:42 MET Received: from cc.FH-Lippe.DE by spock.cc.fh-lippe.de with smtp (Smail3.1.29.1 #2) id m0y41XG-00070FC; Sun, 15 Feb 98 11:42 MET Message-ID: <34E6C6B0.2F02A35B@cc.FH-Lippe.DE> Date: Sun, 15 Feb 1998 11:42:56 +0100 From: "Lars Köller" Organization: Fachhochschule Lippe (Lemgo), Germany X-Mailer: Mozilla 4.04 [en] (Win95; I) MIME-Version: 1.0 To: Lars Koeller CC: Joerg Wunsch , scsi@FreeBSD.ORG, questions@FreeBSD.ORG, deischen@iworks.InterWorks.org, dwhite@gdi.uoregon.edu Subject: Re: SCB and 2940 Problems (with Seagate Python DAT)! References: <199802121231.AA231496715@pollux.hrz.uni-bielefeld.de> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello again! Sorry, but my previous message about the working Seagate Dat stramer was a litte to forwared, I have to revoke it :-((( Using dump seems to be o.k so far, but afbackup, a really nice network backup system produces the same error message as before. You can get the latest version of afbackup from ftp://ftp.zn.ruhr-uni-bochum.de/pub/linux Try it! Sorry again, I will try my best at Seagate again. It looks like the firmware isn't ready for 'tagged command queueing' in all details. Lars Koeller wrote: > > ---------- > > Hello alltogether! > > In reply to J Wunsch who wrote: > > > As Lars Koeller wrote: > > > > > Thank you all for the tips and suggestions. I will try my best at > > > Seagate support! > > Yesterday they send me a firmware upgrade to 5ACB (before 5.AC). It > could be downloaded to the DAT with a DOS bootdisk and an Adaptec > ASPI driver. After the upgrade all problems are gone away. No longer > a SCB timeout or anything else! > > The upgrade can be found on > > ftp://ftp.Uni-Bielefeld.DE/pub/systems/FreeBSD/lkoeller/5acb.exe > > It's a self extracting file DOS executable which contains a README, > the firmware and the download utility for DOS. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 04:30:32 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id EAA29110 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 04:30:32 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from news1.gtn.com (news1.gtn.com [194.77.0.15]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id EAA29103 for ; Sun, 15 Feb 1998 04:30:27 -0800 (PST) (envelope-from andreas@klemm.gtn.com) Received: (from uucp@localhost) by news1.gtn.com (8.8.6/8.8.6) with UUCP id NAA15687; Sun, 15 Feb 1998 13:00:21 +0100 (MET) Received: (from andreas@localhost) by klemm.gtn.com (8.8.8/8.8.7) id MAA22591; Sun, 15 Feb 1998 12:48:55 +0100 (CET) (envelope-from andreas) Message-ID: <19980215124855.47083@klemm.gtn.com> Date: Sun, 15 Feb 1998 12:48:55 +0100 From: Andreas Klemm To: "Justin T. Gibbs" Cc: scsi@FreeBSD.ORG Subject: Re: Amanda 2.4b6 port References: <19980214180849.43287@klemm.gtn.com> <199802150706.AAA16260@narnia.plutotech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <199802150706.AAA16260@narnia.plutotech.com>; from Justin T. Gibbs on Sun, Feb 15, 1998 at 12:06:58AM -0700 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sun, Feb 15, 1998 at 12:06:58AM -0700, Justin T. Gibbs wrote: > In article <19980214224908.62213@klemm.gtn.com> you wrote: > >> p.s. there are still a few issues to look at with the port - --with-fqdn & > >> mmap for a couple. I also am looking at the tape driver PR kern/5735 that > >> Andreas and another fellow have open which is tickled by Amanda. c.t. > > > > Ouch ... I misunderstand something. The TDC 4222 seems to be ok. > > Patrick Hausen told me, that under Linux the drive runs ok with > > amanda. So I think it's a FreeBSD SCSI problem. > > > > Justin, please help ;-) > > Run CAM. Would like to do so, but I build up my complete system based on -current from last weekend including a recompile of all ports. I tried to apply your patches to -current, but I think on one or more files it doesn't apply cleanly. I proceeded with "fingers crossed" ignoring the reject(s). make depend and compile runs well. But during make I get problems. So do you have perhaps a newer snapshot ? Or do you have an idea/possibility to solve the amanda problem ? In the company we have amanda, too, to backup machines. When migrating to -FreeBSD it would be crucial to have a working amanda port. And I wanna kick the linux machines out. The DLD version that is installed currently allows mounting of a dirty filesys. Such a crap ... Andreas /// -- Andreas Klemm powered by ,,symmetric multiprocessor FreeBSD'' To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 08:59:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id IAA24060 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 08:59:09 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id IAA24054 for ; Sun, 15 Feb 1998 08:59:06 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id JAA23855; Sun, 15 Feb 1998 09:58:57 -0700 (MST) Message-Id: <199802151658.JAA23855@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Andreas Klemm cc: "Justin T. Gibbs" , scsi@FreeBSD.ORG Subject: Re: Amanda 2.4b6 port In-reply-to: Your message of "Sun, 15 Feb 1998 10:28:05 +0100." <19980215102805.21437@klemm.gtn.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 15 Feb 1998 09:56:08 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> > Justin, please help ;-) >> Run CAM. > >How available please ? The snapshots have been on ftp.FreeBSD.org for some time now. >-- >Andreas Klemm >powered by ,,symmetric multiprocessor FreeBSD'' -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 09:12:29 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA25771 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 09:12:29 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA25766 for ; Sun, 15 Feb 1998 09:12:27 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id KAA24479; Sun, 15 Feb 1998 10:12:17 -0700 (MST) Message-Id: <199802151712.KAA24479@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Andreas Klemm cc: "Justin T. Gibbs" , scsi@FreeBSD.ORG Subject: Re: Amanda 2.4b6 port In-reply-to: Your message of "Sun, 15 Feb 1998 12:48:55 +0100." <19980215124855.47083@klemm.gtn.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 15 Feb 1998 10:09:29 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> Run CAM. > >Would like to do so, but I build up my complete system based on >-current from last weekend including a recompile of all ports. > >I tried to apply your patches to -current, but I think on one >or more files it doesn't apply cleanly. I proceeded with "fingers >crossed" ignoring the reject(s). This is probably because I added the bus_dma stuff to current a little while after that snapshot. I plan to release another one once I've completed aic7890 support. This should be soon. >So do you have perhaps a newer snapshot ? Or do you have an >idea/possibility to solve the amanda problem? There are lots of problems that CAM fixes that might address your Amanda problem. None of these would be easy to bring into the current SCSI layer. -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 10:39:48 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA05484 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 10:39:48 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from why.whine.com (whine-gw.whine.com [205.150.249.254]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA05461 for ; Sun, 15 Feb 1998 10:39:44 -0800 (PST) (envelope-from andrew@whine.com) Received: from why (why [205.150.249.1]) by why.whine.com (8.8.8/8.7.3) with SMTP id NAA14054 for ; Sun, 15 Feb 1998 13:39:31 -0500 (EST) Date: Sun, 15 Feb 1998 13:39:31 -0500 (EST) From: Andrew Herdman X-Sender: andrew@why To: scsi@FreeBSD.ORG Subject: CRW4260tx with cdrecord (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I recently purchased a new Yamaha CRW4260tx CD rewriteable and CD-R. I am running freebsd-current build on 980205. After sending an e-mail to the author of CDRecord to find out why I was getting: cd0: physio split the request.. cannot proceed He pointed me to dejanews looking for "physio split" which turned up an e-mail for bug kern/5599 which is the exact same problem I am gettting except with the pt0 device. It noted in the bug report that the 971225 snap worked fine, I tried this kernel as well, the volia i can write both CD-R's and CD-RW's. I see there were a few commits to the scsi portion of the sys tree, but can't see anything that would cause this problem, the message is printed from scsi_ioctl.c from function scsistrategy. The commits were done by "eivind". Anyone have any clues? I really want to stick to the snap i'm using now because it's been very stable for me. But I really hate rebooting an old kernel just to write a CD. Thanks Andrew ---------- Forwarded message ---------- Date: Thu, 12 Feb 1998 18:53:59 -0500 (EST) From: Andrew Herdman To: hackers@freebsd.org, schilling@fokus.gmd.de Subject: CRW4260tx with cdrecord I just purchased a new Yamaha CRW4260tx and would like to use cdrecord to write both RW CD's and RO CD's. Problem is i'm getting an odd error. The drive is probed as a scsi cd and comes up as cd0, i have linked /dev/rcd0.ctl to /dev/scgx which the instructions seem to indicate. After doing this, cdrecord is able to access the drive as shown below: # cdrecord -v dev=1,1,0 speed=2 /usr/iso.img Cdrecord release 1.5 Copyright (C) 1995-1997 Jvrg Schilling TOC Type: 1 = CD-ROM scsidev: '1,1,0' scsibus: 1 target: 1 lun: 0 Device type : Removable CD-ROM Version : 2 Response Format: 2 Capabilities : Vendor_info : 'YAMAHA ' Identifikation : 'CRW4260 ' Revision : '1.0e' Device seems to be: Generic mmc CD-RW. Using generic SCSI-3/mmc CD-R driver (mmc_cdr). Driver flags : SWABAUDIO Track 01: data 40 Mb Total size: 46 Mb (04:37.46) = 20810 sectors track info: 00 1A 01 01 00 00 4F 01 00 00 00 00 00 00 00 00 00 05 17 B4 00 00 00 00 00 05 17 B4 Set Starting to write CD at speed 2 in write mode for single session. Last chance to quit, starting real write in 1 seconds. Set Set track info: 00 1A 01 01 00 00 4F 01 00 00 00 00 00 00 00 00 00 05 17 B4 00 00 00 00 00 05 17 B4 Starting new track at sector: 0 cdrecord: Input/output error. write_g1: scsi sendcmd: no error status: 0x0 (GOOD STATUS) resid: 63488 write track data: error after 0 bytes Sense Bytes: 70 00 00 00 00 00 00 0A 00 00 00 00 00 00 00 00 00 00 Writing time: 5.073s Fixating... cdrecord: Undefined error: 0. close session: scsi sendcmd: retryable error status: 0x0 (GOOD STATUS) Fixating time: 0.004s Now, this is the error message that the kernel gives me, and it really concerns me, as I don't see any switches to control data size... cd0: physio split the request.. cannot proceed Has anyone ever run into this problem before and found a way to correct it? I haven't tried CD-R media yet, but the error seems not to be a media issue. Thanks Andrew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 10:55:40 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA08253 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 10:55:40 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from opus.cts.cwu.edu (skynyrd@opus.cts.cwu.edu [198.104.92.71]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA08240 for ; Sun, 15 Feb 1998 10:55:32 -0800 (PST) (envelope-from skynyrd@opus.cts.cwu.edu) Received: from localhost (skynyrd@localhost) by opus.cts.cwu.edu (8.8.8/8.8.8) with SMTP id KAA10208; Sun, 15 Feb 1998 10:55:31 -0800 (PST) (envelope-from skynyrd@opus.cts.cwu.edu) Date: Sun, 15 Feb 1998 10:55:31 -0800 (PST) From: Chris Timmons To: freebsd-scsi@FreeBSD.ORG cc: "Patrick M. Hausen" , John Polstra Subject: Tandberg Tapes & Adaptec In-Reply-To: <199802151304.OAA14194@hugo01.ka.punkt.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org It appears that the problem Patrick is seeing with his Tandberg tape has been around for a while. Visit http://w.dejanews.com/home_ps.shtml and select the OLD button. Then feed the search engine this string: tandberg AND illegal AND tape You'll quickly realize that some other OS's have a problem also. Patrick has pointed out that some linux and solaris OS's revisions work fine. Some people claim firmware has helped them although there is no clear "this is it" answer. Somebody posted a detailed error dump from digital unix' CAM subsystem which might be interesting to the more technically aware scsi hackers here. -Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 12:53:54 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA20946 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 12:53:54 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from bock.salnet.net (d01a8548.dip.cdsnet.net [208.26.133.72]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id MAA20919; Sun, 15 Feb 1998 12:53:50 -0800 (PST) (envelope-from stevel@mail.cdsnet.net) Received: from mail.cdsnet.net (localhost.salnet.net [127.0.0.1]) by bock.salnet.net (8.8.7/8.8.7) with ESMTP id MAA00522; Sun, 15 Feb 1998 12:53:42 -0800 Message-ID: <34E755D6.AED49076@mail.cdsnet.net> Date: Sun, 15 Feb 1998 20:53:42 +0000 From: Steve Logue Organization: nettek LLC X-Mailer: Mozilla 4.04 [en] (X11; U; Linux 2.0.33 i586) MIME-Version: 1.0 To: freebsd-current@FreeBSD.ORG, freebsd-scsi@FreeBSD.ORG Subject: Jaz Drives / Tagged Command Queuing Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org FreeBSD Lists, Due to my own problems as the owner of a Jaz drive, I have gotten word from Iomega that confirms the state of Tagged Command Queuing as the underlying problem. There is an error in all Jaz, and Jaz2 drives prior to BIOS level J.86 that has not shipped yet. Read the following, and make the appropriate corrections to your system present, and future: > Steve, > > I got a very fast response from the hardware engineer (Jaz and Jaz 2 > designer). The problem is this - The Jaz drive does not support > command queing, and revisions older than J.86 do not report it correctly. > For example, when your SCSI adapter says "I'm going to use command > queing" to the Jaz drive, the drive answers "OK, lets go", even though its > not supported. The J.86 drives will now answer "Sorry, command > queing is not supported". Iomega does not have any current plans to > support command queing. > > Thank's for your report, I will continue to forward it to the hardware > engineers. -STEVEl -- --------------------------------------------------------------------- Steve Logue http://home.cdsnet.net/~stevel Systems Integration nettek LLC --------------------------------------------------------------------- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 13:16:32 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA23274 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 13:16:32 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from cabri.obs-besancon.fr (cabri.obs-besancon.fr [193.52.184.3]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id NAA23256 for ; Sun, 15 Feb 1998 13:16:25 -0800 (PST) (envelope-from jmz@cabri.obs-besancon.fr) Received: by cabri.obs-besancon.fr (5.57/Ultrix3.0-C) id AA16525; Sun, 15 Feb 98 22:19:48 +0100 Date: Sun, 15 Feb 98 22:19:48 +0100 Message-Id: <9802152119.AA16525@cabri.obs-besancon.fr> From: Jean-Marc Zucconi To: andrew@whine.com Cc: scsi@FreeBSD.ORG In-Reply-To: (message from Andrew Herdman on Sun, 15 Feb 1998 13:39:31 -0500 (EST)) Subject: Re: CRW4260tx with cdrecord (fwd) X-Mailer: Emacs Mime-Version: 1.0 (generated by tm-edit 7.106) Content-Type: text/plain; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >>>>> Andrew Herdman writes: > I recently purchased a new Yamaha CRW4260tx CD rewriteable and CD-R. > I am running freebsd-current build on 980205. After sending an e-mail > to the author of CDRecord to find out why I was getting: > cd0: physio split the request.. cannot proceed The problem is with kern_physio.c. Revert to revision 1.22. Jean-Marc _____________________________________________________________________________ Jean-Marc Zucconi Observatoire de Besancon F 25010 Besancon cedex PGP Key: finger jmz@cabri.obs-besancon.fr To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 14:00:08 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA28036 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 14:00:08 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from news1.gtn.com (news1.gtn.com [192.109.159.3]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA27987 for ; Sun, 15 Feb 1998 14:00:01 -0800 (PST) (envelope-from andreas@klemm.gtn.com) Received: (from uucp@localhost) by news1.gtn.com (8.8.6/8.8.6) with UUCP id WAA24234; Sun, 15 Feb 1998 22:15:12 +0100 (MET) Received: (from andreas@localhost) by klemm.gtn.com (8.8.8/8.8.7) id VAA14669; Sun, 15 Feb 1998 21:26:44 +0100 (CET) (envelope-from andreas) Message-ID: <19980215212644.24185@klemm.gtn.com> Date: Sun, 15 Feb 1998 21:26:44 +0100 From: Andreas Klemm To: "Justin T. Gibbs" Cc: scsi@FreeBSD.ORG Subject: Re: Amanda 2.4b6 port References: <19980215124855.47083@klemm.gtn.com> <199802151712.KAA24479@pluto.plutotech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <199802151712.KAA24479@pluto.plutotech.com>; from Justin T. Gibbs on Sun, Feb 15, 1998 at 10:09:29AM -0700 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sun, Feb 15, 1998 at 10:09:29AM -0700, Justin T. Gibbs wrote: > >> Run CAM. > > > >Would like to do so, but I build up my complete system based on > >-current from last weekend including a recompile of all ports. > > > >I tried to apply your patches to -current, but I think on one > >or more files it doesn't apply cleanly. I proceeded with "fingers > >crossed" ignoring the reject(s). > > This is probably because I added the bus_dma stuff to current a little > while after that snapshot. I plan to release another one once I've > completed aic7890 support. This should be soon. Fine, thanks, 'til then I'll try a -current of January ! > >So do you have perhaps a newer snapshot ? Or do you have an > >idea/possibility to solve the amanda problem? > > There are lots of problems that CAM fixes that might address your > Amanda problem. None of these would be easy to bring into the > current SCSI layer. Do you think that CAM support could come into 2.2-STABLE if it's working well in -current ? -- Andreas Klemm powered by ,,symmetric multiprocessor FreeBSD'' To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 14:03:20 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA28865 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 14:03:20 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA28859 for ; Sun, 15 Feb 1998 14:03:17 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id PAA04739; Sun, 15 Feb 1998 15:03:10 -0700 (MST) Message-Id: <199802152203.PAA04739@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Andreas Klemm cc: "Justin T. Gibbs" , scsi@FreeBSD.ORG Subject: Re: Amanda 2.4b6 port In-reply-to: Your message of "Sun, 15 Feb 1998 21:26:44 +0100." <19980215212644.24185@klemm.gtn.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 15 Feb 1998 15:00:21 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >Do you think that CAM support could come into 2.2-STABLE if it's >working well in -current ? CAM cannot come into current, let alone stable, until full support for all of the devices the current SCSI code supports is completed. Since I'm the only one working on controller driver support for CAM, this may be a while. I'm always looking for volunteers to commandeer a controller driver or two... >-- >Andreas Klemm >powered by ,,symmetric multiprocessor FreeBSD'' -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 14:58:02 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA06731 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 14:58:02 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from BITS.bris.ac.uk (root@bits.bris.ac.uk [137.222.64.1]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA06700 for ; Sun, 15 Feb 1998 14:57:56 -0800 (PST) (envelope-from stewart@BITS.bris.ac.uk) Received: from BITS.bris.ac.uk (stewart@BITS.bris.ac.uk [137.222.64.1]) by BITS.bris.ac.uk (8.8.8/8.7.3) with SMTP id WAA09609 for ; Sun, 15 Feb 1998 22:58:32 GMT Date: Sun, 15 Feb 1998 22:58:31 +0000 (GMT) From: Stewart Morgan To: freebsd-scsi@FreeBSD.ORG Subject: AIC-7895 SCSI controller Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi there, I've recently purchased a Tyan Thunder 2 (S2696DLUA) motherboard with built in SCSI support... unfortunately, I've been unable to pursuade BSD to recognise the chipset: AIC-7895. My question is: will a driver be available soon? Or, can BSD be fudged to recognise it as a 3940AUW (the motherboard manual says the two are equivalent)? Stewart ------- - Systems-Representative & Joint Systems Administrator BITS - Bristol Information Technology Society University of Bristol Student's Union, Bristol, England E-Mail: stewart@BITS.bris.ac.uk WWW : http://www.BITS.bris.ac.uk/stewart/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sun Feb 15 15:21:04 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA11278 for freebsd-scsi-outgoing; Sun, 15 Feb 1998 15:21:04 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from top.worldcontrol.com (surf52.cruzers.com [205.215.232.52]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA11262 for ; Sun, 15 Feb 1998 15:21:00 -0800 (PST) (envelope-from brian@worldcontrol.com) From: brian@worldcontrol.com Received: (qmail 912 invoked by uid 100); 15 Feb 1998 23:22:26 -0000 Message-ID: <19980215152224.40242@top.worldcontrol.com> Date: Sun, 15 Feb 1998 15:22:24 -0800 To: freebsd-current@FreeBSD.ORG, freebsd-scsi@FreeBSD.ORG Subject: Re: Jaz Drives / Tagged Command Queuing Mail-Followup-To: freebsd-current@freebsd.org, freebsd-scsi@freebsd.org References: <34E755D6.AED49076@mail.cdsnet.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.90.3i In-Reply-To: <34E755D6.AED49076@mail.cdsnet.net>; from Steve Logue on Sun, Feb 15, 1998 at 08:53:42PM +0000 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On %M 0, Steve Logue wrote: > > I got a very fast response from the hardware engineer (Jaz and Jaz 2 > > designer). The problem is this - The Jaz drive does not support > > command queing, and revisions older than J.86 do not report it correctly. > > For example, when your SCSI adapter says "I'm going to use command > > queing" to the Jaz drive, the drive answers "OK, lets go", even though its > > not supported. The J.86 drives will now answer "Sorry, command > > queing is not supported". Iomega does not have any current plans to > > support command queing. I'm not familiar with Iomega products, however, my Syquest Syjet 1.5 at leasts reports that it supports it: ahc0:A:4: refuses WIDE negotiation. Using 8bit transfers ahc0: target 4 Tagged Queuing Device sd1 at scbus0 target 4 lun 0 sd1: type 0 removable SCSI 2 sd1: Direct-Access 1430MB (2929800 512 byte sectors) I have not had any problems using it. A cool aspect of this drive is: 'Minimum Data Transfer Rate: 3.7mb/sec' so you can do AV type stuff with it. -- Brian Litzinger To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Mon Feb 16 09:31:40 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA21738 for freebsd-scsi-outgoing; Mon, 16 Feb 1998 09:31:40 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from gatorade.doctord.com (gatorade.doctord.com [205.230.20.146]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA21733 for ; Mon, 16 Feb 1998 09:31:38 -0800 (PST) (envelope-from schwarz@gatorade.doctord.com) Received: (from schwarz@localhost) by gatorade.doctord.com (8.8.5/8.8.5) id JAA16643; Mon, 16 Feb 1998 09:30:26 -0800 (PST) Message-ID: X-Mailer: XFMail 1.1 [p0] on FreeBSD Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Mon, 16 Feb 1998 09:11:59 -0800 (PST) Organization: Doctor Design From: Steven Schwarz To: Stewart Morgan Subject: RE: AIC-7895 SCSI controller Cc: freebsd-scsi@FreeBSD.ORG Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 15-Feb-98 Stewart Morgan wrote: > Hi there, > > I've recently purchased a Tyan Thunder 2 (S2696DLUA) motherboard with > built in SCSI support... unfortunately, I've been unable to pursuade BSD > to recognise the chipset: AIC-7895. > > My question is: will a driver be available soon? Or, can BSD be fudged > to recognise it as a 3940AUW (the motherboard manual says the two are > equivalent)? For what it's worth, I have this motherboard's onboard SCSI working just fine with a system built from Justin Gibbs' CAM snapshot from 8 December 1997 applied to a -current system from the same date. [Later snaps of the CAM software exist. Check for them at the freebsd ftp site.] I think I understand correctly that enough is different about the AIC-7895 with respect to other earlier 78xx's that small patches to the -stable or -current sources will simply not get you a working AIC-7895. (The strategy employed in writing the 78xx driver does not work on the AIC-7895 because Tyan did not implement the registers used.) So, depending on your tolerance for running non-released kernel code on this system, you may be able to get it going fairly quickly; or you may have to add a supported SCSI card, and bide your time on using the on-board SCSI, til the CAM code is committed and released (which I understand may be more than a little while). sts To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Tue Feb 17 12:06:58 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA09489 for freebsd-scsi-outgoing; Tue, 17 Feb 1998 12:06:58 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from clifford.inch.com (omar@clifford.inch.com [207.240.140.163]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id MAA09442 for ; Tue, 17 Feb 1998 12:06:48 -0800 (PST) (envelope-from omar@clifford.inch.com) Received: (from omar@localhost) by clifford.inch.com (8.8.5/8.8.5) id PAA17552; Tue, 17 Feb 1998 15:01:36 -0500 Message-ID: <19980217150134.49019@clifford.inch.com> Date: Tue, 17 Feb 1998 15:01:34 -0500 From: Omar Thameen To: freebsd-scsi@FreeBSD.ORG Subject: getting oriented with RAID Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.81 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, With DPT's single channel RAID controllers being priced at under $800 retail, we're looking into adding it to several of our systems. I've been reading in the archives and on Simon's website http://simon-shapiro.org), but there are some basic things I'm still not clear on that I'm hoping y'all could explain. 1) From the DPT website, their PM2044UR (single channel PCI) RAID controller supports RAID 0, 1, 5, 0+1, and 0+5. Does this necessarily mean that the FreeBSD driver written by Simon can also do all of the above? Something I read in the archives indicated that you need do use ccd for RAID0. 2) Do the FreeBSD drivers support the PM2044UR card, or do you have to use the higher-end, multi-channel product? 3) Given that you want redundancy and optimal performance, can you increase performance by running RAID0+1, or is it better to just run RAID1? Thanks, Omar To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Tue Feb 17 13:46:20 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA28337 for freebsd-scsi-outgoing; Tue, 17 Feb 1998 13:46:20 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from mail.kcwc.com (h1.kcwc.com [206.139.252.2]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id NAA28282 for ; Tue, 17 Feb 1998 13:46:02 -0800 (PST) (envelope-from curt@kcwc.com) Received: by mail.kcwc.com (NX5.67c/NeXT-2.0-KCWC-1.0) id AA00337; Tue, 17 Feb 98 16:45:54 -0500 Date: Tue, 17 Feb 98 16:45:54 -0500 From: curt@kcwc.com (Curt Welch) Message-Id: <9802172145.AA00337@mail.kcwc.com> Received: by NeXT.Mailer (1.87.1) Received: by NeXT Mailer (1.87.1) To: freebsd-scsi@FreeBSD.ORG Subject: Re: getting oriented with RAID Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > I've been reading in the archives and on Simon's website > http://simon-shapiro.org), but there are some basic things > I'm still not clear on that I'm hoping y'all could > explain. I'll comment on what I know about. > 1) From the DPT website, their PM2044UR (single channel > PCI) RAID controller supports RAID 0, 1, 5, 0+1, and 0+5. > Does this necessarily mean that the FreeBSD driver written > by Simon can also do all of the above? Something I read > in the archives indicated that you need do use ccd for > RAID0. You can do all of those (at least on the PM3334), but as you read, you must use the FreeBSD ccd driver to do the RAID 0 (striping) parts. DPT supports RAID 0 across multiple controllers, which means it must be supported in the driver and not on/in the controller. As far as I can tell, Simon just didn't get around to adding that to the driver. I see no real reason to do so since it's supported by ccd with no problem. ccd is very easy to use and you would have to use it if you wanted to stripe across DPT and non-DPT controllers anyway. If you define a RAID 0 array with the DPT manager, it just shows up as separate drives when you boot instead of as one large drive (like it does with a RAID 1 or RAID 5 array). > 3) Given that you want redundancy and optimal performance, > can you increase performance by running RAID0+1, or is > it better to just run RAID1? I guess you need to explain a little more about what you are asking. I'd say that four 4.5gig drives in a RAID 0+1 config will perform better than two 9gig drives in a RAID 1 config. Both of these configs give you the same overall disk space (9gig), but with the smaller drives striped together, you spread the IO over more spindles (4 vs 2) and therefor you will get better performance. So, in this case 0+1 is better than just 1. But if you have 4 drives which you can leave as two RAID-1 arrays or combine them into one large RAID 0+1 array, then performance won't change much. Curt Welch To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Tue Feb 17 23:25:35 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA07415 for freebsd-scsi-outgoing; Tue, 17 Feb 1998 23:25:35 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA07407 for ; Tue, 17 Feb 1998 23:25:31 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id AAA08524; Wed, 18 Feb 1998 00:25:28 -0700 (MST) Message-Id: <199802180725.AAA08524@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Marty Gordon cc: "Justin T. Gibbs" , Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Tue, 17 Feb 1998 20:55:30 PST." <3.0.3.32.19980217205530.0071563c@mail> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 18 Feb 1998 00:22:38 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >I'm shocked to hear that Western Digital SCSI drives rate really low. All >my illusions (delusions?) are shattered. Who is King of the Mountain for >SCSI-3 drives in terms of reliability and performance? I have a Fast/Wide >controller and I am looking at drives. Without a doubt, IBM. >Thanks, >Marty >mlghome@home.com -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Tue Feb 17 23:57:02 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA11106 for freebsd-scsi-outgoing; Tue, 17 Feb 1998 23:57:02 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id XAA11063 for ; Tue, 17 Feb 1998 23:56:50 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 24627 invoked by uid 1000); 18 Feb 1998 08:02:31 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-020998 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <19980217150134.49019@clifford.inch.com> Date: Wed, 18 Feb 1998 00:02:30 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: Omar Thameen Subject: RE: getting oriented with RAID Cc: freebsd-scsi@FreeBSD.ORG Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 17-Feb-98 Omar Thameen wrote: > Hi, > > With DPT's single channel RAID controllers being priced at under $800 > retail, we're looking into adding it to several of our systems. > > I've been reading in the archives and on Simon's website > http://simon-shapiro.org), but there are some basic things I'm still > not clear on that I'm hoping y'all could explain. > > 1) From the DPT website, their PM2044UR (single channel PCI) RAID > controller supports RAID 0, 1, 5, 0+1, and 0+5. Does this necessarily > mean that the FreeBSD driver written by Simon can also do all of > the above? Something I read in the archives indicated that you > need do use ccd for RAID0. Let me clarify this a bit; DPT always creates RAID-5 arrays by using the firmware on the HBA, and these arrays are confined to one controller (i.e. no more than 7-45 drives, depending on the controller and options). RAID-0 (striping) and RAID-1 (mirroring can be created as either firmware controlled arrays (run dptmggr with /fw0 option), or as in-O/S driver code. Now, with the in-kernel RAID-{0,1} you can even create RAID-0 arrays that are composed of other, redundant arrays. An example says it the best: We have a PM3334UW, with 3 SCSI busses. Each bus has 4 disks, for a total of 12 drives. The drives are all 1GB each. All the drives are setup to be target ID 0-3, This means that our system has drives 0-0, 0-1, 0-2, 0-3, 1-0, 1-1, 1-2, 1-3, 2-0, 2-1, 2-2, 2-3. Using dptmgr, we created 2 RAID-5 arrays, 3 disks in each, 1 RAID-1 array, and 1 disk designated Hot Spare: Array 0: 0-0, 1-0, 2-0 = 2GB usable/visible space Array 1: 0-1, 1-1, 2-1 = 2GB usable/visible space Array 2: 0-2, 1-2, 2-2 = 2GB usable/visible space Array 3: 0-3, 1-3 = 1GB usable/visible space Hot Spare: 2-3 Invisible to the system Now, with my driver, on FreeBSD, this is where it ends for the present time. If, in this example, you cannot live with 3x2GB drives, you have to stripe arrays 0,1,2 together using ccd, and have one ``drive'' with a capacity of 6GB. This array will be very fault tolerant and rather fast (compared to RAID-5 arrays. Actually, you can concatenate Array 3 to the other,s but let's keep it simple. With DOS, Win95, UnixWare, etc. You cna have the DPT create the 6GB super-array. this is called 0+5, or some such. Why can't we do it in FreeBSD? I have not done it yet; To provide this facility, I have to merge a lot of code that will NOT pass ANY FreeBSD coding style. The code literally have to be re-typed. This takes time, and since ccd does the same thing, it has a low priority. > 2) Do the FreeBSD drivers support the PM2044UR card, or do you have > to use the higher-end, multi-channel product? Yes. No. Any PCI card is already in -current. The EISA patches will go in for review later this week. Unlike some other SCSI HBA manufacturers, ALL DPT cards speak the same exact language. So, once the probing and attaching of the driver to the hardware is done, the rest of the driver is unchanged. to be totally honest, some older cards have some minor firmware bugs that has to be worked around. Really old ISA cards, do not have DMA support. These will not work with the driver as it stands today. the change is nimor, on the order of 20-50 lines of code, but it is really a waste of time, as they are out of production for many years. > 3) Given that you want redundancy and optimal performance, can you > increase performance by running RAID0+1, or is it better to just run > RAID1? READ performance of RAID-1 is similar to TAID-0 across two drives. WRITE perfromance is similar to a single drive. READ performance of RAID-5 is similar to RAID-0 of the same number of stripes. WRITE perfromance is much less. On the order of 5-6MB/Sec. The next generation controllers do RAID-5 parity in hardware and have many more MIPS in the processor, so RAID-5 perfromance is greatly enhanced. > > Thanks, > Omar > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-scsi" in the body of the message ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 00:05:20 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA12220 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 00:05:20 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id AAA12211 for ; Wed, 18 Feb 1998 00:05:12 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 3583 invoked by uid 1000); 18 Feb 1998 08:10:54 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-020998 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <9802172145.AA00337@mail.kcwc.com> Date: Wed, 18 Feb 1998 00:10:54 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: (Curt Welch) Subject: Re: getting oriented with RAID Cc: freebsd-scsi@FreeBSD.ORG Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 17-Feb-98 Curt Welch wrote: ... > If you define a RAID 0 array with the DPT manager, it just > shows up as separate drives when you boot instead of as one > large drive (like it does with a RAID 1 or RAID 5 array). Use dptmgr/fw0, or tell dptmgr, during initial installation that your O/S is Linux, or both. I forget which. It will then behave properly. I have several such arrays on nomisd-sendero (the arrays are shared between the two machines). >> 3) Given that you want redundancy and optimal performance, >> can you increase performance by running RAID0+1, or is >> it better to just run RAID1? > > I guess you need to explain a little more about what you > are asking. > > I'd say that four 4.5gig drives in a RAID 0+1 config will > perform better than two 9gig drives in a RAID 1 config. > Both of these configs give you the same overall disk space > (9gig), but with the smaller drives striped together, you > spread the IO over more spindles (4 vs 2) and therefor you > will get better performance. So, in this case 0+1 is better > than just 1. For most random access applications, the more heads the marrier. One of the biggest secrets in RDBMS benchmarking is disk optimization. What you see, almost universally, is that, for high-end systems, the benchmarking engineer rarely uses more than 300MB per drive, and rarely puts more than 4-6 drives on one SCSI bus. Before you jump, Unix ffs filesystems, make, gcc, etc. are a different bird. They typically read large chunks sequentially. OLTP transactions are very small and cache hit ratios deteriorate with database size - fast. Unless you want to do what I used to do for a living for many years (compute all this nonsense), just experiment with drives, busses, stripe sizes, amount of cache, cache utilization, host cache vs. DPT cache, etc. > But if you have 4 drives which you can leave as two RAID-1 > arrays or combine them into one large RAID 0+1 array, then > performance won't change much. > > Curt Welch > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-scsi" in the body of the message ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 02:53:52 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id CAA06455 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 02:53:52 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from ns1.dpt.com (ns1.dpt.com [206.138.241.7]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id CAA06441 for ; Wed, 18 Feb 1998 02:53:42 -0800 (PST) (envelope-from mark@bohica.net) Received: from deathstar.deathstar.dpt.com (deathstar.dpt.com [198.242.63.87]) by ns1.dpt.com (8.7.4/8.7.3) with SMTP id FAA06466 for ; Wed, 18 Feb 1998 05:56:59 -0500 Received: by deathstar.deathstar.dpt.com [198.242.63.87] (NX5.67g/NX3.0M) id AA00457; Wed, 18 Feb 98 05:50:09 -0500 Reply-To: Mark Gregory Salyzyn Message-Id: <9802181050.AA00457@deathstar.deathstar.dpt.com> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.2mach v148) Received: by NeXT.Mailer (1.148) From: Mark Gregory Salyzyn Date: Wed, 18 Feb 98 05:50:06 -0500 To: freebsd-scsi@FreeBSD.ORG Subject: Re: getting oriented with RAID Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In article <9802172145.AA00337@mail.kcwc.com>, you wrote: >> 1) From the DPT website, their PM2044UR (single channel >> PCI) RAID controller supports RAID 0, 1, 5, 0+1, and 0+5. >> Does this necessarily mean that the FreeBSD driver written >> by Simon can also do all of the above? Something I read >> in the archives indicated that you need do use ccd for >> RAID0. RAID support on the PM2044 requires the Caching card to be added ... >You can do all of those (at least on the PM3334), but as >you read, you must use the FreeBSD ccd driver to do the RAID 0 >(striping) parts. DPT supports RAID 0 across multiple >controllers, which means it must be supported in the driver >and not on/in the controller. As far as I can tell, Simon >just didn't get around to adding that to the driver. I see >no real reason to do so since it's supported by ccd with no >problem. ccd is very easy to use and you would have to use >it if you wanted to stripe across DPT and non-DPT controllers >anyway. One reason for adding support is the ability to boot off an array striped across adapters. >If you define a RAID 0 array with the DPT manager, it just >shows up as separate drives when you boot instead of as one >large drive (like it does with a RAID 1 or RAID 5 array). You need to create the array using `other UNIX' selected under storage manager. Another way, with the DOS storage manager, is to run `dptmgr /fw0'. Otherwise the storage manager creates software striped arrays ... Sincerely -- Mark Salyzyn To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 09:28:41 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA19849 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 09:28:41 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from mail3.aracnet.com (root@mail3.aracnet.com [205.159.88.34]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA19765; Wed, 18 Feb 1998 09:28:15 -0800 (PST) (envelope-from beattie@aracnet.com) Received: from shelob.aracnet.com (beattie@shelob.aracnet.com [205.159.88.2]) by mail3.aracnet.com (8.8.5/8.8.5) with ESMTP id JAA20869; Wed, 18 Feb 1998 09:28:00 -0800 Received: from localhost by shelob.aracnet.com (8.8.5) id JAA23052; Wed, 18 Feb 1998 09:28:00 -0800 Date: Wed, 18 Feb 1998 09:27:59 -0800 (PST) From: Brian Beattie To: freebsd-hackers@FreeBSD.ORG, freebsd-scsi@FreeBSD.ORG Subject: looking for soemthing to do. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Please excuse me if this is a duplicate. I'm looking for a project that woudl get me back into the kernel. I've hacjed around in several UNIX kernels going back to V6 but have only looked at the FreeBSD kernel so far. I was thinking about getting an Advansys SCSI controller and writting a driver for that but I also noticed a note indicating that work was needed on the SCSI drivers to support CAM. I have in my posession the following Adaptec controllers 1522, 1542, 2940U. While I could just dive in and start working I woudl really like to coordinate my efforts so as to not duplicate the work of others. I have sent mail to Justin asking about this but have received no reply. On the assumption that he is away or too busy to reply I am asking for input from the mailing list. Anbody got an advice? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 10:00:24 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA26850 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 10:00:24 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from mail3.aracnet.com (michaels@mail3.aracnet.com [205.159.88.34]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA26524; Wed, 18 Feb 1998 09:58:53 -0800 (PST) (envelope-from beattie@aracnet.com) Received: from shelob.aracnet.com (beattie@shelob.aracnet.com [205.159.88.2]) by mail3.aracnet.com (8.8.5/8.8.5) with ESMTP id GAA15136; Wed, 18 Feb 1998 06:54:19 -0800 Received: from localhost by shelob.aracnet.com (8.8.5) id GAA19677; Wed, 18 Feb 1998 06:54:18 -0800 Date: Wed, 18 Feb 1998 06:54:18 -0800 (PST) From: Brian Beattie To: freebsd-hackers@FreeBSD.ORG, freebsd-scsi@FreeBSD.ORG Subject: looking for soemthing to do. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I'm looking for a project that woudl get me back into the kernel. I've hacjed around in several UNIX kernels going back to V6 but have only looked at the FreeBSD kernel so far. I was thinking about getting an Advansys SCSI controller and writting a driver for that but I also noticed a note indicating that work was needed on the SCSI drivers to support CAM. I have in my posession the following Adaptec controllers 1522, 1542, 2940U. While I could just dive in and start working I woudl really like to coordinate my efforts so as to not duplicate the work of others. I have sent mail to Justin asking about this but have received no reply. On the assumption that he is away or too busy to reply I am asking for input from the mailing list. Anbody got an advice? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 10:08:30 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA28546 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 10:08:30 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from thunderdome.plutotech.com (root@thunderdome.plutotech.com [206.168.67.122]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA28397 for ; Wed, 18 Feb 1998 10:07:36 -0800 (PST) (envelope-from ken@panzer.plutotech.com) Received: from panzer.plutotech.com (ken@panzer.plutotech.com [206.168.67.125]) by thunderdome.plutotech.com (8.8.7/8.8.5) with ESMTP id LAA16567; Wed, 18 Feb 1998 11:07:24 -0700 (MST) Received: (from ken@localhost) by panzer.plutotech.com (8.8.8/8.8.5) id LAA12266; Wed, 18 Feb 1998 11:07:22 -0700 (MST) From: "Kenneth D. Merry" Message-Id: <199802181807.LAA12266@panzer.plutotech.com> Subject: Re: looking for soemthing to do. In-Reply-To: from Brian Beattie at "Feb 18, 98 09:27:59 am" To: beattie@aracnet.com (Brian Beattie) Date: Wed, 18 Feb 1998 11:07:22 -0700 (MST) Cc: freebsd-scsi@FreeBSD.ORG, gibbs@pluto.plutotech.com X-Mailer: ELM [version 2.4ME+ PL28s (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Brian Beattie wrote... > Please excuse me if this is a duplicate. > > I'm looking for a project that woudl get me back into the kernel. I've > hacjed around in several UNIX kernels going back to V6 but have only > looked at the FreeBSD kernel so far. I was thinking about getting an > Advansys SCSI controller and writting a driver for that but I also noticed > a note indicating that work was needed on the SCSI drivers to support CAM. > I have in my posession the following Adaptec controllers 1522, 1542, > 2940U. There is already an Advansys driver in CAM, and the Adaptec 2xxx controllers work pretty well. We do however need someone to work on drivers for the 1522 and 1542. Help on that would be much appreciated. (The only SCSI drivers that have been ported to CAM are the Adaptec (ahc), Advansys and NCR drivers...everything else needs work.) > While I could just dive in and start working I woudl really like to > coordinate my efforts so as to not duplicate the work of others. I have > sent mail to Justin asking about this but have received no reply. On the > assumption that he is away or too busy to reply I am asking for input from > the mailing list. He's just busy. If you do want to work on a 1542 or 1522 driver, or any other piece of CAM, contact me via private email and I'll see what I can do to get you setup and running with our development tree. Ken -- Kenneth Merry ken@plutotech.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 11:14:51 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA14030 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 11:14:51 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA13940 for ; Wed, 18 Feb 1998 11:14:13 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y5D2i-000518-00; Wed, 18 Feb 1998 09:12:20 -0800 Date: Wed, 18 Feb 1998 09:12:19 -0800 (PST) From: Tom To: "Justin T. Gibbs" cc: Marty Gordon , Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-Reply-To: <199802180725.AAA08524@pluto.plutotech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 18 Feb 1998, Justin T. Gibbs wrote: > >I'm shocked to hear that Western Digital SCSI drives rate really low. All > >my illusions (delusions?) are shattered. Who is King of the Mountain for > >SCSI-3 drives in terms of reliability and performance? I have a Fast/Wide > >controller and I am looking at drives. > > Without a doubt, IBM. I have some doubts about that. Seagate Barracuda work really well too. 20 drives in 24x7 so far, and no failures. Of course, anything is better than cheap Quantum junk. Of the 5 or so, Quantum drives we have, one is dead and one is "unreliable". > >Thanks, > >Marty > >mlghome@home.com > > -- > Justin Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 11:52:26 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA25900 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 11:52:26 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA25751 for ; Wed, 18 Feb 1998 11:51:38 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 14451 invoked by uid 1000); 18 Feb 1998 19:30:23 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-020998 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199802181850.TAA01578@yedi.iaf.nl> Date: Wed, 18 Feb 1998 11:30:23 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: Wilko Bulte Subject: Re: getting oriented with RAID Cc: freebsd-scsi@FreeBSD.ORG, curt@kcwc.com Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 18-Feb-98 Wilko Bulte wrote: ... > You want to stay on the outside of your ZBR platters to achieve the > highest datarates. My age is showing ... :-) Yes, on modern disks, the number of sectors per platter varies, decreasing towards the spindle. >> Unless you want to do what I used to do for a living for many years >> (compute >> all this nonsense), just experiment with drives, busses, stripe sizes, >> amount of cache, cache utilization, host cache vs. DPT cache, etc. > > Not to forget write-back caching (with battery backup please) Of course. Caching policy has a huge performace impact. but you know, at a certain point it all dies anyway. Consider a recent project with over 1 billion records in the database. Cache hit rate is nil, either way. the only thing that helps there is elevator sorts and deep execution queues. ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 14:02:04 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA25200 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 14:02:04 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from boromir.vpop.net (dns1.vpop.net [206.117.147.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA25163 for ; Wed, 18 Feb 1998 14:01:53 -0800 (PST) (envelope-from mreimer@vpop.net) Received: from vpop.net (bilbo.vpop.net [206.117.152.4]) by boromir.vpop.net (8.8.5/8.8.5) with ESMTP id NAA28933 for ; Wed, 18 Feb 1998 13:35:47 -0800 (PST) Message-ID: <34EB5428.4BE6F8F0@vpop.net> Date: Wed, 18 Feb 1998 13:35:37 -0800 From: Matthew Reimer Organization: VPOP Technologies, Inc. X-Mailer: Mozilla 4.04 [en] (X11; I; FreeBSD 2.2.5-STABLE i386) MIME-Version: 1.0 To: freebsd-scsi@FreeBSD.ORG Subject: Problems with disconnects and 2940UW (Justin?) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I've been having a lot of problems with my 2.2-STABLE system hanging due to SCSI errors. The problem disappears when I turn off "Enable Disconnection" in the Adaptec settings. Might there be a bug in the ahc driver that is manifested when SCSI disconnects/reconnects occur? Someone else I've corresponded with had the same problem that disappeared when he turned them off. The machine is a PII 233MHz, Adaptec 2940UW, one IBM DCAS-34330W 4.3G UW SCSI disk, and 128M RAM. It is running 2.2-STABLE newer than 1 February 1998. The problem was easily reproducible by running 'make world', or by running 'make world' and 'make XFree86' at the same time. The system would hang with errors on the console like these: -------------- sd0(ahc0:0:0): SCB 0x1 - timed out in dataout phase, SCSISIGI == 0xe6 SEQADDR = 0x130 SCSISEQ = 0x12 SSTAT0 = 0x2 SSTAT1 = 0x13 sd0(ahc0:0:0): abort message in message buffer sd0(ahc0:0:0): SCB 0x0 timed out while recovery in progress sd0(ahc0:0:0): SCB 0x1 - timed out in dataout phase, SCSISIGI == 0xf6 SEQADDR = 0x130 SCSISEQ = 0x12 SSTAT0 = 0x2 SSTAT1 = 0x13 sd0(ahc0:0:0): no longer in timeout ahc0: Issued Channel A Bus Reset. 2 SCBs aborted sd0(ahc0:0:0): SCB 0x1 - timed out while idle, LASTPHASE == 0x1, SCSISIGI = 0x0 SEQADDR = 0x166 SCSISEQ = 0x12 SSTAT0 = 0x2 SSTAT1 = 0x0 sd0(ahc0:0:0): Queueing an Abort SCB -------------- dmesg output: ------------- FreeBSD 2.2.5-STABLE #0: Fri Feb 13 15:57:52 PST 1998 root@warp10.smartlink.net:/usr/src/sys/compile/FARAMIR CPU: Pentium Pro (232.67-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0x633 Stepping=3 Features=0x80f9ff real memory = 134217728 (131072K bytes) avail memory = 127397888 (124412K bytes) Probing for devices on PCI bus 0: chip0 rev 2 on pci0:0 chip1 rev 1 on pci0:7:0 chip2 rev 0 on pci0:7:1 ahc0 rev 1 int a irq 10 on pci0:11 ahc0: aic7880 Wide Channel, SCSI Id=7, 16/255 SCBs ahc0: target 0 Tagged Queuing Device (ahc0:0:0): "IBM DCAS-34330W S65A" type 0 fixed SCSI 2 sd0(ahc0:0:0): Direct-Access 4134MB (8467200 512 byte sectors) vga0 rev 211 int a irq 11 on pci0:17 fxp0 rev 2 int a irq 9 on pci0:19 fxp0: Ethernet address 00:a0:c9:8b:c5:d8 Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> ed0 not found at 0x300 sio0 at 0x3f8-0x3ff irq 4 on isa sio0: type 16550A sio1 at 0x2f8-0x2ff irq 3 on isa sio1: type 16550A lpt0 at 0x378-0x37f irq 7 on isa lpt0: Interrupt-driven port lp0: TCP/IP capable interface lpt1 not found at 0xffffffff mse0 not found at 0x23c fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa fdc0: FIFO enabled, 8 bytes threshold fd0: 1.44MB 3.5in npx0 flags 0x1 on motherboard npx0: INT 16 interface -------- The termination is correct, double-, triple- and quadruple-checked, and even forced in the Adaptec settings. I tried three different cables, always hooking the drive to the last connector on the cable, if there were more than one. I repeatedly and carefully checked the jumpers on the drive and motherboard. I eventually turned off every speed-up: Initiate Sync Negotiation = no Initiate Wide Negotiation = no Maximum Synce Transfer Rate = 5MB/s no AHC_TAGENABLE no AHC_SCBPAGING_ENABLE no AHC_ALLOW_MEMIO Even this would fail. But when I set "Enable Disconnection" = no, I can turn on every one of these speed-ups, and no amount of effort can make it crash (simultaneously running make world, make XFree86, and 'find -exec grep' the whole filesystem), even at full Ultra speed with all the kernel options on. Is this a problem with the ahc driver, or something else? Thanks in advance for any answers. Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 14:03:11 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA25408 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 14:03:11 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA25378 for ; Wed, 18 Feb 1998 14:02:59 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id OAA25726; Wed, 18 Feb 1998 14:46:38 -0700 (MST) Message-Id: <199802182146.OAA25726@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Tom cc: Marty Gordon , Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Wed, 18 Feb 1998 09:12:19 PST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 18 Feb 1998 14:43:48 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> Without a doubt, IBM. > > I have some doubts about that. Seagate Barracuda work really well >too. 20 drives in 24x7 so far, and no failures. Many recent Seagate drives are okay, but when I have the choice, I pick IBM over Seagate. This has as much to do with reliability as with how well behaved SCSI protocol wise, the IBM drives are. Their firmware is rock solid and their reliability numbers leave Seagate in the dust. Granted, Pluto has been using mostly Seagate drives in it's Video DDR products for some time now, but there has always been the desire to use IBM instead. Now that IBM is making a strong move to better support the standard retail channel and will guarantee drive allocation to other than IBM internal customers (IBM is it's own biggest customer when it comes to storage products), it looks like this will be possible. The last time I saw the Seagate rep, he was pissing his pants over IBM. 8-) -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 14:35:42 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA01582 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 14:35:42 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from uni4nn.gn.iaf.nl (osmium.gn.iaf.nl [193.67.144.12]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id OAA01492 for ; Wed, 18 Feb 1998 14:35:02 -0800 (PST) (envelope-from wilko@yedi.iaf.nl) Received: by uni4nn.gn.iaf.nl with UUCP id AA09130 (5.67b/IDA-1.5 for freebsd-scsi@FreeBSD.ORG); Wed, 18 Feb 1998 20:11:00 +0100 Received: (from wilko@localhost) by yedi.iaf.nl (8.8.7/8.6.12) id TAA01578; Wed, 18 Feb 1998 19:50:11 +0100 (MET) From: Wilko Bulte Message-Id: <199802181850.TAA01578@yedi.iaf.nl> Subject: Re: getting oriented with RAID In-Reply-To: from Simon Shapiro at "Feb 18, 98 00:10:54 am" To: shimon@simon-shapiro.org Date: Wed, 18 Feb 1998 19:50:11 +0100 (MET) Cc: curt@kcwc.com, freebsd-scsi@FreeBSD.ORG X-Organisation: Private FreeBSD site - Arnhem, The Netherlands X-Pgp-Info: PGP public key at 'finger wilko@freefall.freebsd.org' X-Mailer: ELM [version 2.4ME+ PL32 (25)] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org As Simon Shapiro wrote... > > On 17-Feb-98 Curt Welch wrote: > > ... > > I'd say that four 4.5gig drives in a RAID 0+1 config will > > perform better than two 9gig drives in a RAID 1 config. > > Both of these configs give you the same overall disk space > > (9gig), but with the smaller drives striped together, you > > spread the IO over more spindles (4 vs 2) and therefor you > > will get better performance. So, in this case 0+1 is better > > than just 1. > > For most random access applications, the more heads the marrier. > One of the biggest secrets in RDBMS benchmarking is disk optimization. > What you see, almost universally, is that, for high-end systems, the > benchmarking engineer rarely uses more than 300MB per drive, and rarely > puts more than 4-6 drives on one SCSI bus. You want to stay on the outside of your ZBR platters to achieve the highest datarates. > Unless you want to do what I used to do for a living for many years (compute > all this nonsense), just experiment with drives, busses, stripe sizes, > amount of cache, cache utilization, host cache vs. DPT cache, etc. Not to forget write-back caching (with battery backup please) _ ______________________________________________________________________ | / o / / _ Bulte email: wilko @ yedi.iaf.nl http://www.tcja.nl/~wilko |/|/ / / /( (_) Arnhem, The Netherlands - Do, or do not. There is no 'try' --------------- Support your local daemons: run [Free,Net,Open]BSD Unix -- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:14:24 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA07592 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:14:24 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA07584 for ; Wed, 18 Feb 1998 15:14:12 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y5HsJ-0005GW-00; Wed, 18 Feb 1998 14:21:55 -0800 Date: Wed, 18 Feb 1998 14:21:54 -0800 (PST) From: Tom To: "Justin T. Gibbs" cc: Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-Reply-To: <199802182146.OAA25726@pluto.plutotech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 18 Feb 1998, Justin T. Gibbs wrote: > >> Without a doubt, IBM. > > > > I have some doubts about that. Seagate Barracuda work really well > >too. 20 drives in 24x7 so far, and no failures. > > Many recent Seagate drives are okay, but when I have the choice, I pick > IBM over Seagate. This has as much to do with reliability as with how > well behaved SCSI protocol wise, the IBM drives are. Their firmware is > rock solid and their reliability numbers leave Seagate in the dust. Didn't the IBM DCAS or DORS drive have some firmware bugs in regards to tags? Are you referrring to the newer IBM Ultrastar line? In fact, I just found a message from a DCAS sent to a freebsd list today, that has to disable disconnection to get reliable operation with his 2940 and his DCAS drive. I can't verify that, but I have few problems with 11 Seagate drives hung off a 3940UW. The ahc driver freezes up the system every 6 weeks of heavy 24x7 i/o though. Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:19:05 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA07961 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:19:05 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA07940 for ; Wed, 18 Feb 1998 15:18:59 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id QAA01242; Wed, 18 Feb 1998 16:18:25 -0700 (MST) Message-Id: <199802182318.QAA01242@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Tom cc: "Justin T. Gibbs" , Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Wed, 18 Feb 1998 14:21:54 PST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 18 Feb 1998 16:15:35 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Didn't the IBM DCAS or DORS drive have some firmware bugs in regards to >tags? Are you referrring to the newer IBM Ultrastar line? The problems that have been reported against many IBM drives stem from the fact that the DQUE bit is set on many OEM models. The current SCSI system does not look at the control mode page to see this, attempts to perform tagged queuing operation anyway, and fails miserably. The IBM drives are behaving completely within the bounds of the SCSI spec. CAM handles this "problem" by honoring the DQUE bit assuming that the user knows what they are doing. > In fact, I just found a message from a DCAS sent to a freebsd list >today, that has to disable disconnection to get reliable operation with >his 2940 and his DCAS drive. It is probably the same problem I listed above. Disabling disconnection effectively prevents tags from being used. >I can't verify that, but I have few problems >with 11 Seagate drives hung off a 3940UW. The ahc driver freezes up the >system every 6 weeks of heavy 24x7 i/o though. I'm working on finishing up this 7890 support so that I can release an updated CAM snapshot. Once that happens, you might want to try CAM and see if it addresses this problem. >Tom -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:37:57 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA11487 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:37:57 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA11387 for ; Wed, 18 Feb 1998 15:37:22 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y5ImO-0005JW-00; Wed, 18 Feb 1998 15:19:52 -0800 Date: Wed, 18 Feb 1998 15:19:52 -0800 (PST) From: Tom To: "Justin T. Gibbs" cc: Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-Reply-To: <199802182318.QAA01242@pluto.plutotech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 18 Feb 1998, Justin T. Gibbs wrote: > I'm working on finishing up this 7890 support so that I can release an > updated CAM snapshot. Once that happens, you might want to try CAM > and see if it addresses this problem. But CAM means current. Downgrading production servers to current is not an option. That is why I like the DPT driver for 2.2. The driver/card is quite robust, and I haven't seen it hang ever. I have nearly identical systems (differ only in the amount of RAM and number of disks), one with a DPT and one with a 3940UW. The 3940UW system will reboot complaining about ahc error, every 4 to 6 weeks. The DPT system has never had a problem. Too bad no one in core has imported the DPT driver into -stable yet. > >Tom > > -- > Justin Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:39:59 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA12199 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:39:59 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA12125 for ; Wed, 18 Feb 1998 15:39:44 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y5Iom-0005Jg-00; Wed, 18 Feb 1998 15:22:21 -0800 Date: Wed, 18 Feb 1998 15:22:20 -0800 (PST) From: Tom To: "Justin T. Gibbs" cc: Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-Reply-To: <199802182318.QAA01242@pluto.plutotech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 18 Feb 1998, Justin T. Gibbs wrote: > The problems that have been reported against many IBM drives stem from > the fact that the DQUE bit is set on many OEM models. The current SCSI ... > drives are behaving completely within the bounds of the SCSI spec. CAM > handles this "problem" by honoring the DQUE bit assuming that the user > knows what they are doing. Wait a second... that means that those drives don't support tags at all? Now I'm sure that I don't want IBM drives. Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:46:27 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA14036 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:46:27 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA13934 for ; Wed, 18 Feb 1998 15:45:56 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id QAA02707; Wed, 18 Feb 1998 16:45:45 -0700 (MST) Message-Id: <199802182345.QAA02707@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: Tom cc: "Justin T. Gibbs" , Wee Teck Ng , freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Wed, 18 Feb 1998 15:22:20 PST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 18 Feb 1998 16:42:54 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> drives are behaving completely within the bounds of the SCSI spec. CAM >> handles this "problem" by honoring the DQUE bit assuming that the user >> knows what they are doing. > > Wait a second... that means that those drives don't support tags at all? >Now I'm sure that I don't want IBM drives. I think you should go read the SCSI spec. DQUE means that the user has disabled tagged queuing. If you edit the mode page and turn the bit off, tags work great. >Tom -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:50:46 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA14635 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:50:46 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA14612 for ; Wed, 18 Feb 1998 15:50:35 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 6184 invoked by uid 1000); 18 Feb 1998 23:55:42 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-020998 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Wed, 18 Feb 1998 15:55:42 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: Tom Subject: Re: very slow scsi performance Cc: Wee Teck Ng , freebsd-scsi@FreeBSD.ORG, "Justin T.Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 18-Feb-98 Tom wrote: ... > Too bad no one in core has imported the DPT driver into -stable yet. I will submit a 2.2 patch soon. This week, hopefully. In the meantime, if you need an installation CD, just say so. To me :-) ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 15:52:44 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA14868 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 15:52:44 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA14854 for ; Wed, 18 Feb 1998 15:52:38 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 7813 invoked by uid 1000); 18 Feb 1998 23:57:55 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-020998 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Wed, 18 Feb 1998 15:57:55 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: Tom Subject: Re: very slow scsi performance Cc: Wee Teck Ng , freebsd-scsi@FreeBSD.ORG, "Justin T.Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 18-Feb-98 Tom wrote: > > On Wed, 18 Feb 1998, Justin T. Gibbs wrote: > >> The problems that have been reported against many IBM drives stem from >> the fact that the DQUE bit is set on many OEM models. The current SCSI > ... >> drives are behaving completely within the bounds of the SCSI spec. CAM >> handles this "problem" by honoring the DQUE bit assuming that the user >> knows what they are doing. > > Wait a second... that means that those drives don't support tags at > all? > Now I'm sure that I don't want IBM drives. AFAIK, the DPT firmware handles the IBMs fine. Another DPT advantage (from my point of view); Linking and tagging is done by the firmware. The formware is written by people who have access to a bit more hardware and testing environments than we do. The DPT driver does not attempt tagging. No need to. ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 16:09:20 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id QAA18297 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 16:09:20 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from homer.supersex.com (homer.supersex.com [209.5.1.4]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id QAA18274 for ; Wed, 18 Feb 1998 16:09:15 -0800 (PST) (envelope-from leo@homer.supersex.com) Received: (from leo@localhost) by homer.supersex.com (8.8.8/8.8.5) id TAA05247; Wed, 18 Feb 1998 19:09:37 -0500 (EST) Message-ID: <19980218190936.17836@homer.supersex.com> Date: Wed, 18 Feb 1998 19:09:36 -0500 From: Leo Papandreou To: freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance References: <3.0.3.32.19980217205530.0071563c@mail> <199802180725.AAA08524@pluto.plutotech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.74e In-Reply-To: <199802180725.AAA08524@pluto.plutotech.com>; from Justin T. Gibbs on Wed, Feb 18, 1998 at 12:22:38AM -0700 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, Feb 18, 1998 at 12:22:38AM -0700, Justin T. Gibbs wrote: > >I'm shocked to hear that Western Digital SCSI drives rate really low. All > >my illusions (delusions?) are shattered. Who is King of the Mountain for > >SCSI-3 drives in terms of reliability and performance? I have a Fast/Wide > >controller and I am looking at drives. > > Without a doubt, IBM. I had a Quantum Atlas come through a fire unscathed if not a little bit sooty and crisp around the edges. It smelled pretty bad too. In fact, except for that Quantum, everything else in the system either fried or melted, rendered completely useless. A Seagate Barracuda included. Not only is the Quantum still chugging 2 years, 24x7, later, you now have to hold it pretty damn close to your nose to smell anything at all. I'm so confident in its reliability I dont even bother backing it up any longer. Just joshing, kids [ :-( ] This is a big SCSI-I clunker but I've since installed 3 other Atlas II's alongside old faithful without any hiccups. Granted, the newer 10,000 RPM IBMs spin faster but 7200 rpm is none too shabby. So, if you want to buy some peace of mind based on anecdotal evidence, buy Quantum. > > >Thanks, > >Marty > >mlghome@home.com > > -- > Justin > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-scsi" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 16:32:13 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id QAA22380 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 16:32:13 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from boromir.vpop.net (dns1.vpop.net [206.117.147.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id QAA22371 for ; Wed, 18 Feb 1998 16:32:07 -0800 (PST) (envelope-from mreimer@vpop.net) Received: from vpop.net (bilbo.vpop.net [206.117.152.4]) by boromir.vpop.net (8.8.5/8.8.5) with ESMTP id QAA07562 for ; Wed, 18 Feb 1998 16:32:04 -0800 (PST) Message-ID: <34EB7D7E.B6C78AC6@vpop.net> Date: Wed, 18 Feb 1998 16:31:58 -0800 From: Matthew Reimer Organization: VPOP Technologies, Inc. X-Mailer: Mozilla 4.04 [en] (X11; I; FreeBSD 2.2.5-STABLE i386) MIME-Version: 1.0 To: freebsd-scsi@FreeBSD.ORG Subject: Re: Problems with disconnects and 2940UW (Justin?) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org A couple of other items to add to this scenario: I exchanged the IBM drive with a WD Entrprise 4360 with the same results. The other guy I mentioned that experienced the same problem had 8 Seagate drives in his system; he also reported that the problems cleared up when he turned of disconnects. Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 17:06:27 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA27648 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 17:06:27 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from proxyb1.san.rr.com (proxyb1-atm.san.rr.com [204.210.0.10]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA27639 for ; Wed, 18 Feb 1998 17:06:23 -0800 (PST) (envelope-from Studded@san.rr.com) Received: from san.rr.com (dt050ndd.san.rr.com [204.210.31.221]) by proxyb1.san.rr.com (8.8.7/8.8.8) with ESMTP id RAA10763; Wed, 18 Feb 1998 17:05:21 -0800 (PST) Message-ID: <34EB854D.18B5DF13@san.rr.com> Date: Wed, 18 Feb 1998 17:05:18 -0800 From: Studded Organization: Triborough Bridge and Tunnel Authority X-Mailer: Mozilla 4.04 [en] (X11; I; FreeBSD 2.2.5-STABLE-0218 i386) MIME-Version: 1.0 To: Matthew Reimer CC: freebsd-scsi@FreeBSD.ORG Subject: Re: Problems with disconnects and 2940UW (Justin?) References: <34EB5428.4BE6F8F0@vpop.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Matthew Reimer wrote: > The machine is a PII 233MHz, Adaptec 2940UW, one IBM DCAS-34330W 4.3G UW > SCSI disk, and 128M RAM. It is running 2.2-STABLE newer than 1 February > 1998. I wish I had better news, but I have exactly the same drive and controller and they work like a charm. There is an issue with adaptec drives regarding the Plug and Play SCAM support. It runs a lot better with it enabled. I have the following options set in my adaptec BIOS (version 1.25): Parity Enabled Termination Auto Device options: Sync Yes Transfer rate 40 Disconnection Yes Wide Yes Advanced options: SCAM Enabled Reset Enabled Ultra Enabled Here are my kernel config file options: controller ahc0 options AHC_TAGENABLE options AHC_SCBPAGING_ENABLE options AHC_ALLOW_MEMIO controller scbus0 at ahc0 disk sd0 at scbus0 target 6 <- This is the DCAS disk sd1 at scbus0 target 5 <- This is a Zip drive in case it matters Also someone posted on a related thread that some of the DCAS drives come with the DQUE jumper shorted, you might want to check that too. If you need a BIOS upgrade it's available free from Adaptec, check out their www page and find the link to mail support. Hope this helps, Doug -- *** Chief Operations Officer, DALnet IRC network *** *** Proud operator, designer and maintainer of the world's largest *** Internet Relay Chat server. 5,328 clients and still growing. *** Try spider.dal.net on ports 6662-4 (Powered by FreeBSD) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 17:11:23 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA28877 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 17:11:23 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from allegro.lemis.com (allegro.lemis.com [192.109.197.134]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA28808 for ; Wed, 18 Feb 1998 17:11:04 -0800 (PST) (envelope-from grog@lemis.com) Received: from freebie.lemis.com (freebie.lemis.com [192.109.197.137]) by allegro.lemis.com (8.8.7/8.8.5) with ESMTP id LAA29945 for ; Thu, 19 Feb 1998 11:40:37 +1030 (CST) Received: (from grog@localhost) by freebie.lemis.com (8.8.8/8.8.7) id LAA01932; Thu, 19 Feb 1998 11:40:36 +1030 (CST) (envelope-from grog) Message-ID: <19980219114036.07982@freebie.lemis.com> Date: Thu, 19 Feb 1998 11:40:36 +1030 From: Greg Lehey To: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance References: <199802182146.OAA25726@pluto.plutotech.com> <19980218181631.46512@awesome.us.dell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <19980218181631.46512@awesome.us.dell.com>; from Jerry_Dunham on Wed, Feb 18, 1998 at 06:16:31PM -0600 WWW-Home-Page: http://www.lemis.com/~grog Organization: LEMIS, PO Box 460, Echunga SA 5153, Australia Phone: +61-8-8388-8286 Fax: +61-8-8388-8725 Mobile: +61-41-739-7062 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I forwarded a message to a friend who works at a large computer manufacturer. Here's his reply. He asked that the name of the manufacturer not be revealed. On Wed, 18 February 1998 at 18:16:31 -0600, Jerry_Dunham wrote: > On Wed, Feb 18, 1998 at 02:43:48PM -0700, Justin T. Gibbs wrote: > >>>> Without a doubt, IBM. >>> >>> I have some doubts about that. Seagate Barracuda work really well >>> too. 20 drives in 24x7 so far, and no failures. >> >> Many recent Seagate drives are okay, but when I have the choice, I pick >> IBM over Seagate. This has as much to do with reliability as with how >> well behaved SCSI protocol wise, the IBM drives are. Their firmware is >> rock solid and their reliability numbers leave Seagate in the dust. >> >> Granted, Pluto has been using mostly Seagate drives in it's Video DDR >> products for some time now, but there has always been the desire to use IBM >> instead. Now that IBM is making a strong move to better support the >> standard retail channel and will guarantee drive allocation to other than >> IBM internal customers (IBM is it's own biggest customer when it comes to >> storage products), it looks like this will be possible. The last time I >> saw the Seagate rep, he was pissing his pants over IBM. 8-) > > FWIW, IBM isn't perfect. We've had some serious firmware issues with > them at times, though Seagate is no better, and the 1.6 GB drive we're > now using as our smallest was initially rather fragile (while the 1.4 > and 2.1 were solid). This, of course, only applies to the 2.5-inch > drives, but it's an indication that IBM doesn't walk on water. Based > upon what I see internally (and I don't see everything), I'd choose > IBM first, Fujitsu second, Quantum third, and Seagate fourth, but that > ranking could easily be scrambled if you based it upon individual > products, rather than overall records. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 17:32:08 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA02717 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 17:32:08 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from boromir.vpop.net (dns1.vpop.net [206.117.147.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA02643 for ; Wed, 18 Feb 1998 17:31:47 -0800 (PST) (envelope-from mreimer@vpop.net) Received: from vpop.net (bilbo.vpop.net [206.117.152.4]) by boromir.vpop.net (8.8.5/8.8.5) with ESMTP id RAA10562 for ; Wed, 18 Feb 1998 17:04:28 -0800 (PST) Message-ID: <34EB8517.E628EBAF@vpop.net> Date: Wed, 18 Feb 1998 17:04:23 -0800 From: Matthew Reimer Organization: VPOP Technologies, Inc. X-Mailer: Mozilla 4.04 [en] (X11; I; FreeBSD 2.2.5-STABLE i386) MIME-Version: 1.0 To: freebsd-scsi@FreeBSD.ORG Subject: Re: Problems with disconnects and 2940UW (Justin?) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I just saw Justin's mention of the DQUE bit being set in some IBM OEM drives. It looks like it's not set on either of my drives. Any ideas? Matt # scsi -f /dev/rsd0c -m 10 -P 0 RLEC: 0 Queue Algorithm Modifier: 0 QErr: 0 DQue: 0 EECA: 0 RAENP: 0 UAAENP: 0 EAENP: 0 Ready AEN Holdoff Period: 0 # scsi -f /dev/rsd1c -m 10 -P 0 RLEC: 0 Queue Algorithm Modifier: 1 QErr: 0 DQue: 0 EECA: 0 RAENP: 0 UAAENP: 0 EAENP: 0 Ready AEN Holdoff Period: 0 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 18:21:50 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA12430 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 18:21:50 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id SAA12326 for ; Wed, 18 Feb 1998 18:21:15 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y5LKH-0005Pg-00; Wed, 18 Feb 1998 18:03:01 -0800 Date: Wed, 18 Feb 1998 18:02:55 -0800 (PST) From: Tom To: Greg Lehey cc: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-Reply-To: <19980219114036.07982@freebie.lemis.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Thu, 19 Feb 1998, Greg Lehey wrote: > I forwarded a message to a friend who works at a large computer > manufacturer. Here's his reply. He asked that the name of the > manufacturer not be revealed. ... > > FWIW, IBM isn't perfect. We've had some serious firmware issues with > > them at times, though Seagate is no better, and the 1.6 GB drive we're > > now using as our smallest was initially rather fragile (while the 1.4 > > and 2.1 were solid). This, of course, only applies to the 2.5-inch > > drives, but it's an indication that IBM doesn't walk on water. Based > > upon what I see internally (and I don't see everything), I'd choose > > IBM first, Fujitsu second, Quantum third, and Seagate fourth, but that > > ranking could easily be scrambled if you based it upon individual > > products, rather than overall records. Uhh.. we are talking about SCSI drives here, not IDE. "Large computer manufactures" put cheapo IDE junk in PCs. All the drive manufactures build cheap IDE drives for such "large computer manufacturers". I wouldn't touch either Seagate or IBM IDE drives. Western Digital probably makes the best IDE drives, but that isn't saying much. Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 18:24:45 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA13093 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 18:24:45 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from allegro.lemis.com (allegro.lemis.com [192.109.197.134]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA12991 for ; Wed, 18 Feb 1998 18:24:16 -0800 (PST) (envelope-from grog@lemis.com) Received: from freebie.lemis.com (freebie.lemis.com [192.109.197.137]) by allegro.lemis.com (8.8.7/8.8.5) with ESMTP id MAA00219; Thu, 19 Feb 1998 12:54:10 +1030 (CST) Received: (from grog@localhost) by freebie.lemis.com (8.8.8/8.8.7) id MAA12688; Thu, 19 Feb 1998 12:54:09 +1030 (CST) (envelope-from grog) Message-ID: <19980219125408.65174@freebie.lemis.com> Date: Thu, 19 Feb 1998 12:54:08 +1030 From: Greg Lehey To: Tom Cc: scsi@FreeBSD.ORG, dunham@dunham.org Subject: Re: very slow scsi performance References: <19980219114036.07982@freebie.lemis.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: ; from Tom on Wed, Feb 18, 1998 at 06:02:55PM -0800 WWW-Home-Page: http://www.lemis.com/~grog Organization: LEMIS, PO Box 460, Echunga SA 5153, Australia Phone: +61-8-8388-8286 Fax: +61-8-8388-8725 Mobile: +61-41-739-7062 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 18 February 1998 at 18:02:55 -0800, Tom wrote: > > On Thu, 19 Feb 1998, Greg Lehey wrote: > >> I forwarded a message to a friend who works at a large computer >> manufacturer. Here's his reply. He asked that the name of the >> manufacturer not be revealed. > ... >>> FWIW, IBM isn't perfect. We've had some serious firmware issues with >>> them at times, though Seagate is no better, and the 1.6 GB drive we're >>> now using as our smallest was initially rather fragile (while the 1.4 >>> and 2.1 were solid). This, of course, only applies to the 2.5-inch >>> drives, but it's an indication that IBM doesn't walk on water. Based >>> upon what I see internally (and I don't see everything), I'd choose >>> IBM first, Fujitsu second, Quantum third, and Seagate fourth, but that >>> ranking could easily be scrambled if you based it upon individual >>> products, rather than overall records. > > Uhh.. we are talking about SCSI drives here, not IDE. "Large computer > manufactures" put cheapo IDE junk in PCs. All the drive manufactures > build cheap IDE drives for such "large computer manufacturers". Good point. I didn't think about that. > I wouldn't touch either Seagate or IBM IDE drives. Western Digital > probably makes the best IDE drives, but that isn't saying much. IDE drives are a different matter, but I think you're still being unfair. A lot of IDE drives differ from SCSI drives only by the interface. Greg To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 18:39:56 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA16418 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 18:39:56 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id SAA16348 for ; Wed, 18 Feb 1998 18:39:36 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y5Lcj-0005R3-00; Wed, 18 Feb 1998 18:22:05 -0800 Date: Wed, 18 Feb 1998 18:22:03 -0800 (PST) From: Tom To: Greg Lehey cc: scsi@FreeBSD.ORG, dunham@dunham.org Subject: Re: very slow scsi performance In-Reply-To: <19980219125408.65174@freebie.lemis.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Thu, 19 Feb 1998, Greg Lehey wrote: > IDE drives are a different matter, but I think you're still being > unfair. A lot of IDE drives differ from SCSI drives only by the > interface. That is a myth. The _only_ one I know about is the Quantum Fireball. It is piece of a junk anyhow. I don't know any other drives that are available as either IDE or SCSI. The Seagate Barracuda and IBM Ultrastar drives certainly are not. > Greg Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 18:54:12 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA18955 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 18:54:12 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from gatekeeper.tsc.tdk.com (root@gatekeeper.tsc.tdk.com [207.113.159.21]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA18859 for ; Wed, 18 Feb 1998 18:53:42 -0800 (PST) (envelope-from gdonl@tsc.tdk.com) Received: from sunrise.gv.tsc.tdk.com (root@sunrise.gv.tsc.tdk.com [192.168.241.191]) by gatekeeper.tsc.tdk.com (8.8.8/8.8.8) with ESMTP id SAA07847; Wed, 18 Feb 1998 18:40:55 -0800 (PST) (envelope-from gdonl@tsc.tdk.com) Received: from salsa.gv.tsc.tdk.com (salsa.gv.tsc.tdk.com [192.168.241.194]) by sunrise.gv.tsc.tdk.com (8.8.5/8.8.5) with ESMTP id SAA02132; Wed, 18 Feb 1998 18:40:54 -0800 (PST) Received: (from gdonl@localhost) by salsa.gv.tsc.tdk.com (8.8.5/8.8.5) id SAA13214; Wed, 18 Feb 1998 18:40:52 -0800 (PST) From: Don Lewis Message-Id: <199802190240.SAA13214@salsa.gv.tsc.tdk.com> Date: Wed, 18 Feb 1998 18:40:52 -0800 In-Reply-To: Greg Lehey "Re: very slow scsi performance" (Feb 19, 12:54pm) X-Mailer: Mail User's Shell (7.2.6 alpha(3) 7/19/95) To: Greg Lehey , Tom Subject: Re: very slow scsi performance Cc: scsi@FreeBSD.ORG, dunham@dunham.org Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Feb 19, 12:54pm, Greg Lehey wrote: } Subject: Re: very slow scsi performance } >>> and 2.1 were solid). This, of course, only applies to the 2.5-inch } >>> drives, but it's an indication that IBM doesn't walk on water. Based } >>> upon what I see internally (and I don't see everything), I'd choose } >>> IBM first, Fujitsu second, Quantum third, and Seagate fourth, but that } >>> ranking could easily be scrambled if you based it upon individual } >>> products, rather than overall records. } IDE drives are a different matter, but I think you're still being } unfair. A lot of IDE drives differ from SCSI drives only by the } interface. He's also talking about 2.5-inch drives. So far as I know there aren't any 2.5-inch SCSI drives. --- Truck To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 19:00:41 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id TAA20142 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 19:00:41 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from feral.com (mjacob@[209.54.254.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id TAA20128 for ; Wed, 18 Feb 1998 19:00:34 -0800 (PST) (envelope-from mjacob@feral.com) Received: (from mjacob@localhost) by feral.com (8.8.6/8.8.6) id SAA05112; Wed, 18 Feb 1998 18:58:29 -0800 Date: Wed, 18 Feb 1998 18:58:29 -0800 From: Matthew Jacob Message-Id: <199802190258.SAA05112@feral.com> To: Don.Lewis@tsc.tdk.com, grog@lemis.com, tom@sdf.com Subject: Re: very slow scsi performance Cc: dunham@dunham.org, scsi@FreeBSD.ORG Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >He's also talking about 2.5-inch drives. So far as I know there aren't s Yes. Toshiba makes some. They're what're in the Alpha Multias. About 300MB or so. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 20:39:42 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA01316 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 20:39:42 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from clifford.inch.com (omar@clifford.inch.com [207.240.140.163]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id UAA01279 for ; Wed, 18 Feb 1998 20:39:32 -0800 (PST) (envelope-from omar@clifford.inch.com) Received: (from omar@localhost) by clifford.inch.com (8.8.5/8.8.5) id XAA29651; Wed, 18 Feb 1998 23:33:58 -0500 Message-ID: <19980218233357.20960@clifford.inch.com> Date: Wed, 18 Feb 1998 23:33:57 -0500 From: Omar Thameen To: shimon@simon-shapiro.org Cc: freebsd-scsi@FreeBSD.ORG Subject: Re: getting oriented with RAID References: <19980217150134.49019@clifford.inch.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.81 In-Reply-To: ; from Simon Shapiro on Wed, Feb 18, 1998 at 12:02:30AM -0800 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Thank you for this excellent information. One more question: After you've created RAID arrays with DPT, is it possible to later "grow the array" by adding a drive if you need more space? I'm guessing it might be possible if you've set up RAID-5, but not with 0 or 1. Omar On Wed, Feb 18, 1998 at 12:02:30AM -0800, Simon Shapiro wrote: > > Let me clarify this a bit; > > DPT always creates RAID-5 arrays by using the firmware on the HBA, and > these arrays are confined to one controller (i.e. no more than 7-45 drives, > depending on the controller and options). > > RAID-0 (striping) and RAID-1 (mirroring can be created as either firmware > controlled arrays (run dptmggr with /fw0 option), or as in-O/S driver code. > > Now, with the in-kernel RAID-{0,1} you can even create RAID-0 arrays that > are composed of other, redundant arrays. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 20:49:20 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA02862 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 20:49:20 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from silvia.HIP.Berkeley.EDU (ala-ca34-51.ix.netcom.com [207.93.143.179]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id UAA02857 for ; Wed, 18 Feb 1998 20:49:16 -0800 (PST) (envelope-from asami@vader.cs.berkeley.edu) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.8/8.6.9) id UAA15764; Wed, 18 Feb 1998 20:48:57 -0800 (PST) Date: Wed, 18 Feb 1998 20:48:57 -0800 (PST) Message-Id: <199802190448.UAA15764@silvia.HIP.Berkeley.EDU> To: gibbs@plutotech.com CC: tom@sdf.com, KILLSPAM%mlghome@home.com, weeteck@eecs.umich.edu, freebsd-scsi@FreeBSD.ORG In-reply-to: <199802182146.OAA25726@pluto.plutotech.com> (gibbs@plutotech.com) Subject: Re: very slow scsi performance From: asami@FreeBSD.ORG (Satoshi Asami) Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org * Many recent Seagate drives are okay, but when I have the choice, I pick * IBM over Seagate. This has as much to do with reliability as with how * well behaved SCSI protocol wise, the IBM drives are. Their firmware is * rock solid and their reliability numbers leave Seagate in the dust. I don't know about that. The IBM DCHS ("Scorpion") drives have been nothing but problems for us, and need a lot of schmoozing to get them to work under pressure without locking up. (And I'm sure you know that too, because it's you who helped me get them run. :) As far as manufacturing quality (not firmware) goes, IBM drives are pretty good. We've had ~400 in our systems for about a year and only one has died outright so far. Satoshi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 20:56:41 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA04244 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 20:56:41 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from clifford.inch.com (omar@clifford.inch.com [207.240.140.163]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id UAA04201 for ; Wed, 18 Feb 1998 20:56:20 -0800 (PST) (envelope-from omar@clifford.inch.com) Received: (from omar@localhost) by clifford.inch.com (8.8.5/8.8.5) id XAA29753; Wed, 18 Feb 1998 23:50:59 -0500 Message-ID: <19980218235059.26727@clifford.inch.com> Date: Wed, 18 Feb 1998 23:50:59 -0500 From: Omar Thameen To: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.81 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org While we're on the subject of preferred drives, I should throw in our $.02. We're an ISP, so between news and heavily hit servers, we see some pretty good usage. Back before we settled on Seagate Barracudas, we had an unusual number of problems with Quantum drives, mostly the Gran Prix's. Some of them seem to run fine, but we had 3 of maybe 8-10 go bad, and they all went down due to "stiction" after powering them down. Maybe that has improved with the newer models. We haven't had one of the Barracudas go bad yet, though one is now giving us sd0(bt0:0:0): MEDIUM ERROR info:0x3ba93c asc:11,0 Unrecovered read error field replaceable unit: ea sks:80,b , FAILURE Where can I get more info about these types of errors? We've had about 15-20 of the Barracudas in production for about a year or so, and a dozen 5200 RPM SCSI Connor/Seagates in our news machine (to keep the cost down). At the risk of stating the obvious, there's one thing that can give you cheap insurance: fans. We add 2 fans to every server, one blowing directly on the drives, and one blowing out. Omar On Wed, Feb 18, 1998 at 06:22:03PM -0800, Tom wrote: > > That is a myth. The _only_ one I know about is the Quantum Fireball. > It is piece of a junk anyhow. I don't know any other drives that are > available as either IDE or SCSI. The Seagate Barracuda and IBM Ultrastar > drives certainly are not. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 21:02:37 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA04839 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 21:02:37 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from relay.nuxi.com (nuxi.cs.ucdavis.edu [128.120.56.38]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id VAA04828 for ; Wed, 18 Feb 1998 21:02:29 -0800 (PST) (envelope-from obrien@NUXI.com) Received: (from obrien@localhost) by relay.nuxi.com (8.8.7/8.6.12) id FAA28458; Thu, 19 Feb 1998 05:02:27 GMT Message-ID: <19980218210227.50467@nuxi.com> Date: Wed, 18 Feb 1998 21:02:27 -0800 From: "David E. O'Brien" To: freebsd-scsi@FreeBSD.ORG Cc: tom@sdf.com, KILLSPAM%mlghome@home.com, weeteck@eecs.umich.edu Subject: Re: very slow scsi performance Reply-To: obrien@NUXI.com References: <199802182146.OAA25726@pluto.plutotech.com> <199802190448.UAA15764@silvia.HIP.Berkeley.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88 In-Reply-To: <199802190448.UAA15764@silvia.HIP.Berkeley.EDU>; from Satoshi Asami on Wed, Feb 18, 1998 at 08:48:57PM -0800 X-Warning: Mutt Bites! X-Operating-System: FreeBSD 2.2.5-STABLE Organization: The NUXI *BSD group X-PGP-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Keyid: 34F9F9D5 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > I don't know about that. The IBM DCHS ("Scorpion") drives have been I've seen people mention a lot of IBM models, but nobody has commented on the IBM DFHS drives (they may be OEM only). Does anyone have any experience with them? -- -- David (obrien@NUXI.ucdavis.edu -or- obrien@FreeBSD.org) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 21:38:17 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA08683 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 21:38:17 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id VAA08676 for ; Wed, 18 Feb 1998 21:38:14 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 12043 invoked by uid 1000); 19 Feb 1998 05:46:27 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-021598 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <19980218233357.20960@clifford.inch.com> Date: Wed, 18 Feb 1998 21:46:27 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: Omar Thameen Subject: Re: getting oriented with RAID Cc: freebsd-scsi@FreeBSD.ORG Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 19-Feb-98 Omar Thameen wrote: > Thank you for this excellent information. One more question: > > After you've created RAID arrays with DPT, is it possible to > later "grow the array" by adding a drive if you need more space? > I'm guessing it might be possible if you've set up RAID-5, but not > with 0 or 1. Yes. I do not remember the details. Shrinking an array is a different story :-) > > Omar > > On Wed, Feb 18, 1998 at 12:02:30AM -0800, Simon Shapiro wrote: >> >> Let me clarify this a bit; >> >> DPT always creates RAID-5 arrays by using the firmware on the HBA, and >> these arrays are confined to one controller (i.e. no more than 7-45 >> drives, >> depending on the controller and options). >> >> RAID-0 (striping) and RAID-1 (mirroring can be created as either >> firmware >> controlled arrays (run dptmggr with /fw0 option), or as in-O/S driver >> code. >> >> Now, with the in-kernel RAID-{0,1} you can even create RAID-0 arrays >> that >> are composed of other, redundant arrays. ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 22:09:22 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA12570 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 22:09:22 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA12564; Wed, 18 Feb 1998 22:09:19 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id XAA18516; Wed, 18 Feb 1998 23:09:15 -0700 (MST) Message-Id: <199802190609.XAA18516@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: asami@FreeBSD.ORG (Satoshi Asami) cc: gibbs@plutotech.com, tom@sdf.com, KILLSPAM%mlghome@home.com, weeteck@eecs.umich.edu, freebsd-scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Wed, 18 Feb 1998 20:48:57 PST." <199802190448.UAA15764@silvia.HIP.Berkeley.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 18 Feb 1998 23:06:24 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > * Many recent Seagate drives are okay, but when I have the choice, I pick > * IBM over Seagate. This has as much to do with reliability as with how > * well behaved SCSI protocol wise, the IBM drives are. Their firmware is > * rock solid and their reliability numbers leave Seagate in the dust. > >I don't know about that. The IBM DCHS ("Scorpion") drives have been >nothing but problems for us, and need a lot of schmoozing to get them >to work under pressure without locking up. (And I'm sure you know >that too, because it's you who helped me get them run. :) Haven't you been running "Beta" drives? Did you ever look into getting updated firmware or contacting technical support about the reset problems? >Satoshi -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Wed Feb 18 23:29:08 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA21995 for freebsd-scsi-outgoing; Wed, 18 Feb 1998 23:29:08 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from silvia.HIP.Berkeley.EDU (ala-ca34-51.ix.netcom.com [207.93.143.179]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA21964 for ; Wed, 18 Feb 1998 23:29:03 -0800 (PST) (envelope-from asami@vader.cs.berkeley.edu) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.8/8.6.9) id WAA16259; Wed, 18 Feb 1998 22:33:49 -0800 (PST) Date: Wed, 18 Feb 1998 22:33:49 -0800 (PST) Message-Id: <199802190633.WAA16259@silvia.HIP.Berkeley.EDU> To: gibbs@plutotech.com CC: gibbs@plutotech.com, tom@sdf.com, KILLSPAM%mlghome@home.com, weeteck@eecs.umich.edu, freebsd-scsi@FreeBSD.ORG In-reply-to: <199802190609.XAA18516@pluto.plutotech.com> (gibbs@plutotech.com) Subject: Re: very slow scsi performance From: asami@FreeBSD.ORG (Satoshi Asami) Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org * Haven't you been running "Beta" drives? Did you ever look into getting * updated firmware or contacting technical support about the reset problems? These drives are not Beta, we won't buy 400 of Beta drives. And yes, we are planning to actually go down there to talk with the people when our current rush of projects are over (in a couple of weeks). Satoshi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 01:01:44 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id BAA04943 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 01:01:44 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from relay.nuxi.com (nuxi.cs.ucdavis.edu [128.120.56.38]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id BAA04935 for ; Thu, 19 Feb 1998 01:01:42 -0800 (PST) (envelope-from obrien@dragon.nuxi.com) Received: from dragon.nuxi.com (d60-090.leach.ucdavis.edu [169.237.60.90]) by relay.nuxi.com (8.8.7/8.6.12) with ESMTP id BAA04150; Thu, 19 Feb 1998 01:01:41 -0800 (PST) Received: (from obrien@localhost) by dragon.nuxi.com (8.8.8/8.7.3) id JAA27871; Thu, 19 Feb 1998 09:01:10 GMT Message-ID: <19980219010110.38244@nuxi.com> Date: Thu, 19 Feb 1998 01:01:10 -0800 From: "David O'Brien" To: Matthew Reimer Cc: freebsd-scsi@FreeBSD.ORG Subject: Re: Problems with disconnects and 2940UW (Justin?) Reply-To: obrien@NUXI.com References: <34EB8517.E628EBAF@vpop.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88 In-Reply-To: <34EB8517.E628EBAF@vpop.net>; from Matthew Reimer on Wed, Feb 18, 1998 at 05:04:23PM -0800 X-Warning: Mutt Bites! X-Operating-System: FreeBSD 2.2.5-STABLE Organization: The NUXI *BSD group X-PGP-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Keyid: 34F9F9D5 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > # scsi -f /dev/rsd0c -m 10 -P 0 > Queue Algorithm Modifier: 0 > > # scsi -f /dev/rsd1c -m 10 -P 0 > Queue Algorithm Modifier: 1 What's the implication of the difference in the two drives? -- -- David (obrien@NUXI.com -or- obrien@FreeBSD.org) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 07:45:49 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA04934 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 07:45:49 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from mail.kcwc.com (h1.kcwc.com [206.139.252.2]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id HAA04913 for ; Thu, 19 Feb 1998 07:45:43 -0800 (PST) (envelope-from curt@kcwc.com) Received: by mail.kcwc.com (NX5.67c/NeXT-2.0-KCWC-1.0) id AA02302; Thu, 19 Feb 98 10:45:41 -0500 Date: Thu, 19 Feb 98 10:45:41 -0500 From: curt@kcwc.com (Curt Welch) Message-Id: <9802191545.AA02302@mail.kcwc.com> Received: by NeXT.Mailer (1.87.1) Received: by NeXT Mailer (1.87.1) To: freebsd-scsi@FreeBSD.ORG Subject: Re: getting oriented with RAID Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > On 19-Feb-98 Omar Thameen wrote: > > Thank you for this excellent information. One more question: > > > > After you've created RAID arrays with DPT, is it possible to > > later "grow the array" by adding a drive if you need more space? > > I'm guessing it might be possible if you've set up RAID-5, but not > > with 0 or 1. > > Yes. I do not remember the details. Shrinking an array is a different > story :-) But if you build a Unix File system on that array, there's no way that I know of to grow the file system even if you can grow the underlying array. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 09:12:45 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA15664 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 09:12:45 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from boromir.vpop.net (dns1.vpop.net [206.117.147.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA15624 for ; Thu, 19 Feb 1998 09:12:39 -0800 (PST) (envelope-from mreimer@vpop.net) Received: from vpop.net (bilbo.vpop.net [206.117.152.4]) by boromir.vpop.net (8.8.5/8.8.5) with ESMTP id JAA07714; Thu, 19 Feb 1998 09:12:30 -0800 (PST) Message-ID: <34EC6812.A2381C1D@vpop.net> Date: Thu, 19 Feb 1998 09:12:50 -0800 From: Matthew Reimer Organization: VPOP Technologies, Inc. X-Mailer: Mozilla 4.04 [en] (X11; I; FreeBSD 2.2.5-STABLE i386) MIME-Version: 1.0 To: obrien@NUXI.com CC: freebsd-scsi@FreeBSD.ORG Subject: Re: Problems with disconnects and 2940UW (Justin?) References: <34EB8517.E628EBAF@vpop.net> <19980219010110.38244@nuxi.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org David O'Brien wrote: > > > # scsi -f /dev/rsd0c -m 10 -P 0 > > Queue Algorithm Modifier: 0 > > > > # scsi -f /dev/rsd1c -m 10 -P 0 > > Queue Algorithm Modifier: 1 > > What's the implication of the difference in the two drives? None, I think, because I did the testing with only one of the drives in the machine at a time, and it broke with either drive in place. Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 10:44:17 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA29250 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 10:44:17 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from sendero.simon-shapiro.org (sendero-fxp0.Simon-Shapiro.ORG [206.190.148.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id KAA29169 for ; Thu, 19 Feb 1998 10:44:08 -0800 (PST) (envelope-from shimon@sendero-fxp0.simon-shapiro.org) Received: (qmail 24235 invoked by uid 1000); 19 Feb 1998 18:52:19 -0000 Message-ID: X-Mailer: XFMail 1.3-alpha-021598 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199802191814.TAA01167@yedi.iaf.nl> Date: Thu, 19 Feb 1998 10:52:19 -0800 (PST) Reply-To: shimon@simon-shapiro.org Organization: The Simon Shapiro Foundation From: Simon Shapiro To: Wilko Bulte Subject: Re: getting oriented with RAID Cc: curt@kcwc.com, freebsd-scsi@FreeBSD.ORG Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 19-Feb-98 Wilko Bulte wrote: > Ach. Ultrix-11 on PDP11/34 with 2x RK05, now that's ancient ;-) Now that is OLD :-) ... > For single stream reads etc cache is pretty useless. But you want WB > cache > anyway to avoid RAID5 writehole pittfalls. So, e.g. data has been updated > on disk but the corresponding parity block never made it to disk when the > power went out. It is worse than that. For a single, streaming read, the cache actually represents an overhed. this is how you get a 2940 to perform ``better'' than a DPT 3334 :-) On the other extreame, where the dataset is huge, extreamly fragmented and totally random, caching losses it. Consider a multi-terrabyte database with a random access to 512byte blocks... For RAID-5 a cache is almost always useful. An HBA cache is generally more useful on the WRITE side, thn on the READ side (which the O/S tnds to have), but Mark is welcome to contradict me here :-) ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 11:35:05 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA09525 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 11:35:05 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from uni4nn.gn.iaf.nl (osmium.gn.iaf.nl [193.67.144.12]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA09503 for ; Thu, 19 Feb 1998 11:34:59 -0800 (PST) (envelope-from wilko@yedi.iaf.nl) Received: by uni4nn.gn.iaf.nl with UUCP id AA24460 (5.67b/IDA-1.5 for freebsd-scsi@FreeBSD.ORG); Thu, 19 Feb 1998 19:35:34 +0100 Received: (from wilko@localhost) by yedi.iaf.nl (8.8.7/8.6.12) id SAA01007; Thu, 19 Feb 1998 18:54:55 +0100 (MET) From: Wilko Bulte Message-Id: <199802191754.SAA01007@yedi.iaf.nl> Subject: Re: getting oriented with RAID In-Reply-To: <9802191545.AA02302@mail.kcwc.com> from Curt Welch at "Feb 19, 98 10:45:41 am" To: curt@kcwc.com (Curt Welch) Date: Thu, 19 Feb 1998 18:54:55 +0100 (MET) Cc: freebsd-scsi@FreeBSD.ORG X-Organisation: Private FreeBSD site - Arnhem, The Netherlands X-Pgp-Info: PGP public key at 'finger wilko@freefall.freebsd.org' X-Mailer: ELM [version 2.4ME+ PL32 (25)] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org As Curt Welch wrote... > > On 19-Feb-98 Omar Thameen wrote: > > > Thank you for this excellent information. One more question: > > > > > > > After you've created RAID arrays with DPT, is it possible to > > > later "grow the array" by adding a drive if you need more space? > > > I'm guessing it might be possible if you've set up RAID-5, but not > > > with 0 or 1. > > > > > Yes. I do not remember the details. Shrinking an array is a > different > > story :-) > > But if you build a Unix File system on that array, there's no > way that I know of to grow the file system even if you can grow > the underlying array. I think (I'm not a FS expert) that you could potentially come up with a special hack to newfs to add cylindergroups to an existing FS. And update things like the superblocks etc of course. Real freaks(TM) would want to do that with the FS mounted ;-) Wilko _ ______________________________________________________________________ | / o / / _ Bulte email: wilko @ yedi.iaf.nl http://www.tcja.nl/~wilko |/|/ / / /( (_) Arnhem, The Netherlands - Do, or do not. There is no 'try' --------------- Support your local daemons: run [Free,Net,Open]BSD Unix -- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 11:35:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA09539 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 11:35:09 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from uni4nn.gn.iaf.nl (osmium.gn.iaf.nl [193.67.144.12]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA09524 for ; Thu, 19 Feb 1998 11:35:05 -0800 (PST) (envelope-from wilko@yedi.iaf.nl) Received: by uni4nn.gn.iaf.nl with UUCP id AA24477 (5.67b/IDA-1.5 for freebsd-scsi@FreeBSD.ORG); Thu, 19 Feb 1998 19:35:37 +0100 Received: (from wilko@localhost) by yedi.iaf.nl (8.8.7/8.6.12) id TAA01167; Thu, 19 Feb 1998 19:14:52 +0100 (MET) From: Wilko Bulte Message-Id: <199802191814.TAA01167@yedi.iaf.nl> Subject: Re: getting oriented with RAID In-Reply-To: from Simon Shapiro at "Feb 18, 98 11:30:23 am" To: shimon@simon-shapiro.org Date: Thu, 19 Feb 1998 19:14:52 +0100 (MET) Cc: freebsd-scsi@FreeBSD.ORG, curt@kcwc.com X-Organisation: Private FreeBSD site - Arnhem, The Netherlands X-Pgp-Info: PGP public key at 'finger wilko@freefall.freebsd.org' X-Mailer: ELM [version 2.4ME+ PL32 (25)] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org As Simon Shapiro wrote... > > On 18-Feb-98 Wilko Bulte wrote: > > ... > > > You want to stay on the outside of your ZBR platters to achieve the > > highest datarates. > > My age is showing ... :-) Yes, on modern disks, the number of sectors per > platter varies, decreasing towards the spindle. Ach. Ultrix-11 on PDP11/34 with 2x RK05, now that's ancient ;-) > >> all this nonsense), just experiment with drives, busses, stripe sizes, > >> amount of cache, cache utilization, host cache vs. DPT cache, etc. > > > > Not to forget write-back caching (with battery backup please) > > Of course. Caching policy has a huge performace impact. but you know, at > a certain point it all dies anyway. Consider a recent project with over 1 > billion records in the database. Cache hit rate is nil, either way. the > only thing that helps there is elevator sorts and deep execution queues. For single stream reads etc cache is pretty useless. But you want WB cache anyway to avoid RAID5 writehole pittfalls. So, e.g. data has been updated on disk but the corresponding parity block never made it to disk when the power went out. W/ _ ______________________________________________________________________ | / o / / _ Bulte email: wilko @ yedi.iaf.nl http://www.tcja.nl/~wilko |/|/ / / /( (_) Arnhem, The Netherlands - Do, or do not. There is no 'try' --------------- Support your local daemons: run [Free,Net,Open]BSD Unix -- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Thu Feb 19 11:35:28 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA09596 for freebsd-scsi-outgoing; Thu, 19 Feb 1998 11:35:28 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from uni4nn.gn.iaf.nl (osmium.gn.iaf.nl [193.67.144.12]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA09582 for ; Thu, 19 Feb 1998 11:35:20 -0800 (PST) (envelope-from wilko@yedi.iaf.nl) Received: by uni4nn.gn.iaf.nl with UUCP id AA24432 (5.67b/IDA-1.5 for scsi@FreeBSD.ORG); Thu, 19 Feb 1998 19:35:18 +0100 Received: (from wilko@localhost) by yedi.iaf.nl (8.8.7/8.6.12) id TAA01120; Thu, 19 Feb 1998 19:08:40 +0100 (MET) From: Wilko Bulte Message-Id: <199802191808.TAA01120@yedi.iaf.nl> Subject: Re: very slow scsi performance In-Reply-To: <199802190258.SAA05112@feral.com> from Matthew Jacob at "Feb 18, 98 06:58:29 pm" To: mjacob@feral.com (Matthew Jacob) Date: Thu, 19 Feb 1998 19:08:40 +0100 (MET) Cc: Don.Lewis@tsc.tdk.com, grog@lemis.com, tom@sdf.com, dunham@dunham.org, scsi@FreeBSD.ORG X-Organisation: Private FreeBSD site - Arnhem, The Netherlands X-Pgp-Info: PGP public key at 'finger wilko@freefall.freebsd.org' X-Mailer: ELM [version 2.4ME+ PL32 (25)] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org As Matthew Jacob wrote... > >He's also talking about 2.5-inch drives. So far as I know there aren't > s > Yes. Toshiba makes some. They're what're in the Alpha Multias. > About 300MB or so. Yep, tiny thing. There is an obvious problem with them: they use an exotic header connector pin spacing for the SCSI bus cable. Wilko _ ______________________________________________________________________ | / o / / _ Bulte email: wilko @ yedi.iaf.nl http://www.tcja.nl/~wilko |/|/ / / /( (_) Arnhem, The Netherlands - Do, or do not. There is no 'try' --------------- Support your local daemons: run [Free,Net,Open]BSD Unix -- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Fri Feb 20 21:28:46 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA00208 for freebsd-scsi-outgoing; Fri, 20 Feb 1998 21:28:46 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id VAA00198 for ; Fri, 20 Feb 1998 21:28:43 -0800 (PST) (envelope-from gibbs@narnia.plutotech.com) Received: (from gibbs@localhost) by narnia.plutotech.com (8.8.8/8.7.3) id WAA27892; Fri, 20 Feb 1998 22:25:42 -0700 (MST) Date: Fri, 20 Feb 1998 22:25:42 -0700 (MST) From: "Justin T. Gibbs" Message-Id: <199802210525.WAA27892@narnia.plutotech.com> To: Leo Papandreou cc: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance Newsgroups: pluto.freebsd.scsi In-Reply-To: <3.0.3.32.19980217205530.0071563c@mail> <199802180725.AAA08524@pluto.plutotech.com> <19980218190936.17836@homer.supersex.com> User-Agent: tin/pre-1.4-971204 (UNIX) (FreeBSD/3.0-CURRENT (i386)) Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > This is a big SCSI-I clunker but I've since installed 3 other Atlas > II's alongside old faithful without any hiccups. Granted, the newer > 10,000 RPM IBMs spin faster but 7200 rpm is none too shabby. The Atlas I did pretty well. The Atlas II has a nasty problem with the queue full condition that can render tagged queuing useless unless your queuing algorithm is smart. I'm hopeful that they will fix it for the Atlas III, but even if they don't, CAM knows how to deal with this particular type of problem. -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Fri Feb 20 22:57:58 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA06369 for freebsd-scsi-outgoing; Fri, 20 Feb 1998 22:57:58 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA06364 for ; Fri, 20 Feb 1998 22:57:57 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id XAA19810; Fri, 20 Feb 1998 23:57:42 -0700 (MST) Message-Id: <199802210657.XAA19810@pluto.plutotech.com> To: shimon@simon-shapiro.org cc: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance Newsgroups: pluto.freebsd.scsi In-Reply-To: Date: Fri, 20 Feb 1998 23:54:50 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > AFAIK, the DPT firmware handles the IBMs fine. Another DPT advantage (from > my point of view); Linking and tagging is done by the firmware. The > formware is written by people who have access to a bit more hardware and > testing environments than we do. The DPT driver does not attempt tagging. > No need to. Giving applications the ability to control the tag types is very powerful. It gives you explicit control over transaction ordering. The fact that the DPT cards export an interface to control the tag type is good, but the fact that the controller driver does not export the ability to set it is not. I'm sure that when the DPT driver is integrated into the CAM framework, it will the application specified tag type. > Sincerely Yours, > > Simon Shapiro > Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Fri Feb 20 23:01:00 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA06823 for freebsd-scsi-outgoing; Fri, 20 Feb 1998 23:01:00 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA06816 for ; Fri, 20 Feb 1998 23:00:58 -0800 (PST) (envelope-from gibbs@narnia.plutotech.com) Received: (from gibbs@localhost) by narnia.plutotech.com (8.8.8/8.7.3) id XAA28088; Fri, 20 Feb 1998 23:58:01 -0700 (MST) Date: Fri, 20 Feb 1998 23:58:01 -0700 (MST) From: "Justin T. Gibbs" Message-Id: <199802210658.XAA28088@narnia.plutotech.com> To: Tom cc: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance Newsgroups: pluto.freebsd.scsi In-Reply-To: User-Agent: tin/pre-1.4-971204 (UNIX) (FreeBSD/3.0-CURRENT (i386)) Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In article you wrote: > > On Wed, 18 Feb 1998, Justin T. Gibbs wrote: > >> I'm working on finishing up this 7890 support so that I can release an >> updated CAM snapshot. Once that happens, you might want to try CAM >> and see if it addresses this problem. > > But CAM means current. Downgrading production servers to current is not > an option. CAM works for wcarchive and many of Satoshi's "production" systems. It is possible to port CAM to 2.2 as well since that is how wcarchive is using it. -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Fri Feb 20 23:42:08 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA11402 for freebsd-scsi-outgoing; Fri, 20 Feb 1998 23:42:08 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA11397 for ; Fri, 20 Feb 1998 23:42:06 -0800 (PST) (envelope-from root@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id XAA19276; Fri, 20 Feb 1998 23:40:39 -0800 (PST) Message-Id: <199802210740.XAA19276@implode.root.com> To: "Justin T. Gibbs" cc: Tom , scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Fri, 20 Feb 1998 23:58:01 MST." <199802210658.XAA28088@narnia.plutotech.com> From: David Greenman Reply-To: dg@root.com Date: Fri, 20 Feb 1998 23:40:38 -0800 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> But CAM means current. Downgrading production servers to current is not >> an option. > >CAM works for wcarchive and many of Satoshi's "production" systems. It >is possible to port CAM to 2.2 as well since that is how wcarchive is >using it. I should mention that what I'm using on there has been quite stable as well. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sat Feb 21 08:31:18 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id IAA11319 for freebsd-scsi-outgoing; Sat, 21 Feb 1998 08:31:18 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from ha1.rdc1.az.home.com (siteadm@ha1.rdc1.az.home.com [24.1.240.66]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id IAA11301 for ; Sat, 21 Feb 1998 08:31:16 -0800 (PST) (envelope-from straka@home.com) Received: from straka.motorvation.com ([24.1.209.47]) by ha1.rdc1.az.home.com (Netscape Mail Server v2.02) with SMTP id AAA279; Sat, 21 Feb 1998 08:31:13 -0800 Message-ID: <34EF018E.41C67EA6@home.com> Date: Sat, 21 Feb 1998 09:32:14 -0700 From: "Richard S. Straka" X-Mailer: Mozilla 3.01Gold (X11; I; FreeBSD 3.0-971208-SNAP i386) MIME-Version: 1.0 To: "Justin T. Gibbs" CC: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance References: <199802210525.WAA27892@narnia.plutotech.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Justin T. Gibbs wrote: > > > This is a big SCSI-I clunker but I've since installed 3 other Atlas > > II's alongside old faithful without any hiccups. Granted, the newer > > 10,000 RPM IBMs spin faster but 7200 rpm is none too shabby. > > The Atlas I did pretty well. The Atlas II has a nasty problem with > the queue full condition that can render tagged queuing useless > unless your queuing algorithm is smart. I'm hopeful that they will > fix it for the Atlas III, but even if they don't, CAM knows how to > deal with this particular type of problem. > > -- > Justin > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-scsi" in the body of the message With my Symbios 875 card, I was getting scsi errors indicating the tag queue was full on my Atlas 2 drive. After reading the Atlas 2 manual, I found that the WCE (Writeback Cache Enable) bit was set by default. With this bit set, the drive returns a GOOD status and a TASK COMPLETE after fetching the data and storing it in cache memory. If this bit is not set, GOOD status and TASK COMPLETE is returned only after the data is committed to the media. With the WCE bit set, it looks like the NCR driver kept sending more TAGS until the cache memory filled up. When I set the WCE bit to 0, the scsi errors went away and as long as the number of tags is set high enough (>16) performance did not seem to be adversely effected according to bonnie benchmark runs. regards, Richard Straka To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sat Feb 21 10:05:26 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA21638 for freebsd-scsi-outgoing; Sat, 21 Feb 1998 10:05:26 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA21623 for ; Sat, 21 Feb 1998 10:05:23 -0800 (PST) (envelope-from gibbs@plutotech.com) Received: from narnia.plutotech.com (narnia.plutotech.com [206.168.67.130]) by pluto.plutotech.com (8.8.7/8.8.5) with ESMTP id LAA17699; Sat, 21 Feb 1998 11:05:20 -0700 (MST) Message-Id: <199802211805.LAA17699@pluto.plutotech.com> X-Mailer: exmh version 2.0.1 12/23/97 To: "Richard S. Straka" cc: "Justin T. Gibbs" , scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-reply-to: Your message of "Sat, 21 Feb 1998 09:32:14 MST." <34EF018E.41C67EA6@home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 21 Feb 1998 11:02:27 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> The Atlas I did pretty well. The Atlas II has a nasty problem with >> the queue full condition that can render tagged queuing useless >> unless your queuing algorithm is smart. I'm hopeful that they will >> fix it for the Atlas III, but even if they don't, CAM knows how to >> deal with this particular type of problem. >> >> -- >> Justin > >With my Symbios 875 card, I was getting scsi errors indicating the >tag queue was full on my Atlas 2 drive. After reading the Atlas 2 >manual, I found that the WCE (Writeback Cache Enable) bit was set by >default. With this bit set, the drive returns a GOOD status and a >TASK COMPLETE after fetching the data and storing it in cache memory. >If this bit is not set, GOOD status and TASK COMPLETE is returned only >after the data is committed to the media. Yes. That is exactly how the cache is supposed to work and it should have no effect on the number of outstanding transactions the controller can maintain to the drive. >With the WCE bit set, it looks like the NCR driver kept sending more >TAGS until the cache memory filled up. I would expect the cache to fill up and that this might postpone future write requests, but it should not cause future write requests to fail. No other drive I've tested has this behavior. I assume these "other drives" use separate data structures to keep track of open transactions to the initiator, and cache transactions that are waiting to hit the disk. This means that you can allocate a static amount of memory for transactions the initiator knows about and regardless of how your cache fills up, always service up to a static number of initiator transactions at a time. >When I set the WCE bit to 0, >the scsi errors went away and as long as the number of tags is set high >enough (>16) performance did not seem to be adversely effected >according to bonnie benchmark runs. Not using the write cache really hurts your write latency. You shouldn't have to turn off the cache to make a drive with a coherent cache work properly. This is simply a drive with poorly designed firmware. BTW. CAM is much more aggressive in the way it manages it's queues and tags. It will happily queue up to 64 transactions (this is the current default, but it can easily be set higher) to a drive and dynamically adjust that number depending on how the drive responds. With the Atlas II, a simplistic algorithm will cause the OS to drop down as low as 2 transactions if you hit the drive with the right I/O pattern (say a restore from another disk). To combat this, CAM allows you to specify a minimum tag count (currently set at 24 for Atlas IIs) so that even in the face of temporary queue shortages, you can still keep the transaction count at something reasonable. Of course, the OS still pauses until at least a single transaction completes before sending more when a queue full occurs, but seeing queue fulls at all wastes SCSI bus bandwidth. >regards, > >Richard Straka > -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sat Feb 21 15:22:01 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA05865 for freebsd-scsi-outgoing; Sat, 21 Feb 1998 15:22:01 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from shell.futuresouth.com (shell.futuresouth.com [207.141.254.20]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA05845 for ; Sat, 21 Feb 1998 15:21:56 -0800 (PST) (envelope-from tim@shell.futuresouth.com) Received: (from tim@localhost) by shell.futuresouth.com (8.8.8/8.8.8) id RAA10745; Sat, 21 Feb 1998 17:21:43 -0600 (CST) Message-ID: <19980221172142.33725@futuresouth.com> Date: Sat, 21 Feb 1998 17:21:42 -0600 From: Tim Tsai To: "Justin T. Gibbs" Cc: scsi@FreeBSD.ORG Subject: Re: very slow scsi performance References: <34EF018E.41C67EA6@home.com> <199802211805.LAA17699@pluto.plutotech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88 In-Reply-To: <199802211805.LAA17699@pluto.plutotech.com>; from Justin T. Gibbs on Sat, Feb 21, 1998 at 11:02:27AM -0700 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > that number depending on how the drive responds. With the Atlas II, a > simplistic algorithm will cause the OS to drop down as low as 2 > transactions if you hit the drive with the right I/O pattern (say a restore > from another disk). As somebody who has a number of Atlas II's, does anybody know if Quantum has a fix for this or are we stuck with it? I do not see a firmware section on Quantum's website (not that I really want to do that to a production drive anyway). More reasons to keep buying IBM's I guess. Tim To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sat Feb 21 15:41:27 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA09002 for freebsd-scsi-outgoing; Sat, 21 Feb 1998 15:41:27 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from misery.sdf.com (misery.sdf.com [204.244.213.32]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id PAA08968 for ; Sat, 21 Feb 1998 15:41:25 -0800 (PST) (envelope-from tom@sdf.com) Received: from tom by misery.sdf.com with smtp (Exim 1.73 #1) id 0y6OGp-0000Ws-00; Sat, 21 Feb 1998 15:23:47 -0800 Date: Sat, 21 Feb 1998 15:23:45 -0800 (PST) From: Tom To: Tim Tsai cc: "Justin T. Gibbs" , scsi@FreeBSD.ORG Subject: Re: very slow scsi performance In-Reply-To: <19980221172142.33725@futuresouth.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sat, 21 Feb 1998, Tim Tsai wrote: > > that number depending on how the drive responds. With the Atlas II, a > > simplistic algorithm will cause the OS to drop down as low as 2 > > transactions if you hit the drive with the right I/O pattern (say a restore > > from another disk). > > As somebody who has a number of Atlas II's, does anybody know if Quantum > has a fix for this or are we stuck with it? I do not see a firmware > section on Quantum's website (not that I really want to do that to a > production drive anyway). ftp://ftp.quantum.com/Firmware/ has some stuff. > More reasons to keep buying IBM's I guess. More reasons to keep buying Seagate Barracuda's too... Not that you could buy new Atlas IIs anymore. There were replaced by the Atlas IIIs quite a while ago. > Tim Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message From owner-freebsd-scsi Sat Feb 21 18:05:53 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA27445 for freebsd-scsi-outgoing; Sat, 21 Feb 1998 18:05:53 -0800 (PST) (envelope-from owner-freebsd-scsi@FreeBSD.ORG) Received: from relay.nuxi.com (nuxi.cs.ucdavis.edu [128.120.56.38]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA27440 for ; Sat, 21 Feb 1998 18:05:49 -0800 (PST) (envelope-from obrien@dragon.nuxi.com) Received: from dragon.nuxi.com (d60-090.leach.ucdavis.edu [169.237.60.90]) by relay.nuxi.com (8.8.7/8.6.12) with ESMTP id SAA02376; Sat, 21 Feb 1998 18:05:31 -0800 (PST) Received: (from obrien@localhost) by dragon.nuxi.com (8.8.8/8.7.3) id CAA02867; Sun, 22 Feb 1998 02:05:39 GMT Message-ID: <19980221180539.25431@nuxi.com> Date: Sat, 21 Feb 1998 18:05:39 -0800 From: "David O'Brien" To: dg@root.com Cc: "Justin T. Gibbs" , scsi@FreeBSD.ORG Subject: Re: very slow scsi performance Reply-To: obrien@NUXI.com References: <199802210658.XAA28088@narnia.plutotech.com> <199802210740.XAA19276@implode.root.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88 In-Reply-To: <199802210740.XAA19276@implode.root.com>; from David Greenman on Fri, Feb 20, 1998 at 11:40:38PM -0800 X-Warning: Mutt Bites! X-Operating-System: FreeBSD 2.2.5-STABLE Organization: The NUXI *BSD group X-PGP-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Keyid: 34F9F9D5 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > >> But CAM means current. Downgrading production servers to current is not > >> an option. > > > >CAM works for wcarchive and many of Satoshi's "production" systems. It > >is possible to port CAM to 2.2 as well since that is how wcarchive is > >using it. > > I should mention that what I'm using on there has been quite stable as > well. Any diffs available? -- -- David (obrien@NUXI.com -or- obrien@FreeBSD.org) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message