From owner-freebsd-scsi Sun Oct 6 00:03:47 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id AAA20073 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 00:03:47 -0700 (PDT) Received: from freefall.freebsd.org (localhost.cdrom.com [127.0.0.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id AAA20066; Sun, 6 Oct 1996 00:03:45 -0700 (PDT) Message-Id: <199610060703.AAA20066@freefall.freebsd.org> To: Brett Glass cc: scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? In-reply-to: Your message of "Sun, 06 Oct 1996 00:44:26 MDT." <199610060644.AAA09800@lariat.lariat.org> Date: Sun, 06 Oct 1996 00:03:44 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >Should I just copy the new aic7770.c into /sys/i386/isa and recompile? Or >run the entire -stable kernel against my existing 2.1.5 install? (I'm not >sure how to do it, but I assume the latter would involve saving my kernel >config file, nuking all of /sys, un-tarring some large file into /sys to >replace the kernel source, and running through the usual kernel-building >process....) > >--Brett Have you ever used CVS? Do you have disk space for a CVS tree? If the answer to at least the second question is "yes", you should probably be CVSup'ing the CVS tree and using a checked out copy of stable for your server. If you can't do that, copy the following files (once you see my commit mail for the stable branch) into your source tree and recompile your kernel: sys/i386/scsi/aic7xxx.c sys/i386/scsi/aic7xxx.h sys/dev/aic7xxx/* sys/i386/eisa/aic7770.c sys/pci/aic7870.c sys/scsi/scsi_message.h /* New file */ This *should* compile under 2.1.5, but no guarantees. -- Justin T. Gibbs =========================================== FreeBSD: Turning PCs into workstations =========================================== From owner-freebsd-scsi Sun Oct 6 21:56:24 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id VAA10375 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 21:56:24 -0700 (PDT) Received: from lserver.infoworld.com (lserver.infoworld.com [192.216.48.4]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id VAA10368; Sun, 6 Oct 1996 21:56:21 -0700 (PDT) From: BRETT_GLASS@infoworld.com Received: from ccgate.infoworld.com (ccgate.infoworld.com [192.216.49.101]) by lserver.infoworld.com (8.7.5/8.7.3/GNAC-GW-1.2) with SMTP id VAA00838; Sun, 6 Oct 1996 21:56:18 -0700 (PDT) Received: from ccMail by ccgate.infoworld.com (SMTPLINK V2.11) id AA844663979; Sun, 06 Oct 96 19:32:59 PST Date: Sun, 06 Oct 96 19:32:59 PST Message-Id: <9609068446.AA844663979@ccgate.infoworld.com> To: "Justin T. Gibbs" , brett@lariat.org Cc: scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Have you ever used CVS? 'fraid I never have. Most of my UNIX programming has consisted of monolithic utility programs that I've done all myself.... No real need for version control. I don't know what it would involve to create a CVS tree, but it sounds like overkill for one patch. I admit to being blissfully ignorant of the overall kernel architecture of FreeBSD. (Much of the knowledge seems to be folklore; I've not seen anything written that really explains it. I'd love to attend a seminar on this, if one was given.) I've successfully made small patches to drivers without really understanding the "big picture" (for instance, the one that turns off "green mode" on a Seagate IDE drive), but don't yet know enough to write a NEW driver from scratch. > If you can't do that, copy the following files (once you see my commit > mail for the stable branch) into your source tree and recompile your > kernel: > > sys/i386/scsi/aic7xxx.c > sys/i386/scsi/aic7xxx.h > sys/dev/aic7xxx/* > sys/i386/eisa/aic7770.c > sys/pci/aic7870.c > sys/scsi/scsi_message.h /* New file */ Hmmm.... I've never gotten "commit mail," so I guess I'm not on that list. But I wonder if the organization of the files has changed. 2.1.0 doesn't have a sys/i386/eisa/aic7770.c; the ahc code seems to be in sys/isa. Will the new code "fit?" --Brett From owner-freebsd-scsi Sun Oct 6 22:12:23 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA11103 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 22:12:23 -0700 (PDT) Received: from freefall.freebsd.org (localhost.cdrom.com [127.0.0.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA11097; Sun, 6 Oct 1996 22:12:21 -0700 (PDT) Message-Id: <199610070512.WAA11097@freefall.freebsd.org> To: BRETT_GLASS@infoworld.com cc: brett@lariat.org, scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? In-reply-to: Your message of "Sun, 06 Oct 1996 19:32:59 PST." <9609068446.AA844663979@ccgate.infoworld.com> Date: Sun, 06 Oct 1996 22:12:21 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> Have you ever used CVS? > >'fraid I never have. Most of my UNIX programming has consisted of >monolithic utility programs that I've done all myself.... No real need for >version control. I don't know what it would involve to create a CVS tree, >but it sounds like overkill for one patch. Once your setup, you have total control over what patches come into your "critical" server, and what doesn't. Its actually quite easy to setup: grab CVSup from freefall.FreeBSD.org:/pub/CVSup and follow its directions for pulling down the CVS tree. setenv CVSROOT /dir/where/you/put/cvs/tree cd /usr /*It doesn't have to be in /usr, but mine is*/ rm -rf src /*Kill your old src tree since CVS will *complain that its in the way */ cvs co -rRELENG_2_1_0 src You now have a checked out copy of stable. You can update your cvs tree at will and whenever you want to pull something in, you do something like: cd /usr/src cvs update -rRELENG_2_1_0 If you only want to update a few files, specify them to the update command and that's all that will come in. >I admit to being blissfully ignorant of the overall kernel architecture of >FreeBSD. (Much of the knowledge seems to be folklore; I've not seen >anything written that really explains it. I'd love to attend a seminar on >this, if one was given.) Go to one of the 4.4BSD seminars at Usenix or other *NIX conventions. >I've successfully made small patches to drivers >without really understanding the "big picture" (for instance, the one that >turns off "green mode" on a Seagate IDE drive), but don't yet know enough >to write a NEW driver from scratch. There aren't any "how to" documents yet although there will be one for the SCSI system before USENIX rolls around. >> If you can't do that, copy the following files (once you see my commit >> mail for the stable branch) into your source tree and recompile your >> kernel: >> >> sys/i386/scsi/aic7xxx.c >> sys/i386/scsi/aic7xxx.h >> sys/dev/aic7xxx/* >> sys/i386/eisa/aic7770.c >> sys/pci/aic7870.c >> sys/scsi/scsi_message.h /* New file */ > >Hmmm.... I've never gotten "commit mail," so I guess I'm not on that list. >But I wonder if the organization of the files has changed. 2.1.0 doesn't >have a sys/i386/eisa/aic7770.c; the ahc code seems to be in sys/isa. Will >the new code "fit?" You're running 2.1.0? Ouch. That driver was totally buggy. I thought you were running 2.1.5R. If you pull down the source for 2.1.5R, you can do a "make world" upgrade just by recompiling the whole system. >--Brett -- Justin T. Gibbs =========================================== FreeBSD: Turning PCs into workstations =========================================== From owner-freebsd-scsi Sun Oct 6 22:35:21 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA12852 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 22:35:21 -0700 (PDT) Received: from lariat.lariat.org ([129.72.251.2]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA12837; Sun, 6 Oct 1996 22:35:08 -0700 (PDT) Received: (from brett@localhost) by lariat.lariat.org (8.8.Alpha.4/8.8.Alpha.4) id XAA21049; Sun, 6 Oct 1996 23:33:26 -0600 (MDT) Date: Sun, 6 Oct 1996 23:33:26 -0600 (MDT) From: Brett Glass Message-Id: <199610070533.XAA21049@lariat.lariat.org> To: BRETT_GLASS@infoworld.com, gibbs@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? Cc: brett@lariat.org, scsi@freefall.freebsd.org Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > You're running 2.1.0? Ouch. That driver was totally buggy. Oops.... I am actually running 2.1.5R on that machine. I have another machine here with 2.1.0R (kernel sources only) that I checked to see the general layout of the files. (2.1.0R did NOT run with the Adaptec EISA SCSI board, so I guess it was indeed buggy.) Is there any way to bring in just the kernel using CVS? Or is that advisable? --Brett From owner-freebsd-scsi Sun Oct 6 22:41:22 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA13708 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 22:41:22 -0700 (PDT) Received: from freefall.freebsd.org (localhost.cdrom.com [127.0.0.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA13695; Sun, 6 Oct 1996 22:41:19 -0700 (PDT) Message-Id: <199610070541.WAA13695@freefall.freebsd.org> To: Brett Glass cc: BRETT_GLASS@infoworld.com, scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? In-reply-to: Your message of "Sun, 06 Oct 1996 23:33:26 MDT." <199610070533.XAA21049@lariat.lariat.org> Date: Sun, 06 Oct 1996 22:41:18 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> You're running 2.1.0? Ouch. That driver was totally buggy. > >Oops.... I am actually running 2.1.5R on that machine. I have another >machine here with 2.1.0R (kernel sources only) that I checked to see >the general layout of the files. (2.1.0R did NOT run with the Adaptec >EISA SCSI board, so I guess it was indeed buggy.) Yup. >Is there any way to bring in just the kernel using CVS? Or is that >advisable? > >--Brett You can, but some of your userland utilities may not work (ps, w, mountd, etc). You'll definitely be able to at least boot single user. cd /usr/src rm -rf sys cvs update -rRELENG_2_1_0 sys cd sys/i386/conf config MYKERNEL cd ../../compile/MYKERNEL make make install sync reboot -- Justin T. Gibbs =========================================== FreeBSD: Turning PCs into workstations =========================================== From owner-freebsd-scsi Sun Oct 6 22:50:11 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA14666 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 22:50:11 -0700 (PDT) Received: from lariat.lariat.org ([129.72.251.2]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA14654; Sun, 6 Oct 1996 22:50:05 -0700 (PDT) Received: (from brett@localhost) by lariat.lariat.org (8.8.Alpha.4/8.8.Alpha.4) id XAA21231; Sun, 6 Oct 1996 23:49:46 -0600 (MDT) Date: Sun, 6 Oct 1996 23:49:46 -0600 (MDT) From: Brett Glass Message-Id: <199610070549.XAA21231@lariat.lariat.org> To: brett@lariat.org, gibbs@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? Cc: BRETT_GLASS@infoworld.com, scsi@freefall.freebsd.org Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > You can, but some of your userland utilities may not work (ps, w, mountd, > etc). You'll definitely be able to at least boot single user. Hmmm. Could be painful. But I don't want to wipe out configuration I've done on the system either. I've changed rc.serial, crontab, inetd.conf, namedb, the latest version of sendmail, all of the utilities I've added to /usr/local/bin, etc. How could I upgrade without losing all of this? --Brett From owner-freebsd-scsi Sun Oct 6 22:57:10 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA15153 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 22:57:10 -0700 (PDT) Received: from freefall.freebsd.org (localhost.cdrom.com [127.0.0.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA15145; Sun, 6 Oct 1996 22:57:04 -0700 (PDT) Message-Id: <199610070557.WAA15145@freefall.freebsd.org> To: Brett Glass cc: BRETT_GLASS@infoworld.com, scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? In-reply-to: Your message of "Sun, 06 Oct 1996 23:49:46 MDT." <199610070549.XAA21231@lariat.lariat.org> Date: Sun, 06 Oct 1996 22:57:04 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> You can, but some of your userland utilities may not work (ps, w, mountd, >> etc). You'll definitely be able to at least boot single user. > >Hmmm. Could be painful. But I don't want to wipe out configuration I've done >on the system either. I've changed rc.serial, crontab, inetd.conf, namedb, >the latest version of sendmail, all of the utilities I've added to >/usr/local/bin, etc. How could I upgrade without losing all of this? > >--Brett Why do you think that upgrading would cause you to lose all this? FreeBSD will not touch /usr/local/bin unless you install a port. It will also not touch /etc unless you explicitly ask it to by going into /usr/src/etc and doing a make install. As for sendmail, yes, doing a make world will clober sendmail, but probably with the same version you're using right now (unless you decided to be experimental and use 8.8). -- Justin T. Gibbs =========================================== FreeBSD: Turning PCs into workstations =========================================== From owner-freebsd-scsi Sun Oct 6 23:19:49 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id XAA16912 for freebsd-scsi-outgoing; Sun, 6 Oct 1996 23:19:49 -0700 (PDT) Received: from lariat.lariat.org ([129.72.251.2]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id XAA16907; Sun, 6 Oct 1996 23:19:42 -0700 (PDT) Received: (from brett@localhost) by lariat.lariat.org (8.8.Alpha.4/8.8.Alpha.4) id AAA21478; Mon, 7 Oct 1996 00:19:17 -0600 (MDT) Date: Mon, 7 Oct 1996 00:19:17 -0600 (MDT) From: Brett Glass Message-Id: <199610070619.AAA21478@lariat.lariat.org> To: brett@lariat.org, gibbs@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? Cc: BRETT_GLASS@infoworld.com, scsi@freefall.freebsd.org Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Why do you think that upgrading would cause you to lose all this? Only because some of the docs (I forget whether it was in one of the FAQs or in the Handbook) said so. They said there was currently no good way of doing an automatic upgrade, and that the manual process was painful and could cause you to lose things. If there's a way around it, I just need to know what it is.... --Brett From owner-freebsd-scsi Mon Oct 7 06:55:25 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id GAA27763 for freebsd-scsi-outgoing; Mon, 7 Oct 1996 06:55:25 -0700 (PDT) Received: from freefall.freebsd.org (localhost.cdrom.com [127.0.0.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id GAA27757; Mon, 7 Oct 1996 06:55:22 -0700 (PDT) Message-Id: <199610071355.GAA27757@freefall.freebsd.org> To: Brett Glass cc: BRETT_GLASS@infoworld.com, scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? In-reply-to: Your message of "Mon, 07 Oct 1996 00:19:17 MDT." <199610070619.AAA21478@lariat.lariat.org> Date: Mon, 07 Oct 1996 06:55:22 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> Why do you think that upgrading would cause you to lose all this? > >Only because some of the docs (I forget whether it was in one of the FAQs >or in the Handbook) said so. They said there was currently no good way >of doing an automatic upgrade, and that the manual process was painful >and could cause you to lose things. If there's a way around it, >I just need to know what it is.... > >--Brett Its called, "cd /usr/src; make world". Since you are already running 2.1.5R, this will work just fine for you. Nothing has changed in /etc since the release of 2.1.5R on the "stable" branch. -- Justin T. Gibbs =========================================== FreeBSD: Turning PCs into workstations =========================================== From owner-freebsd-scsi Mon Oct 7 09:09:09 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id JAA08403 for freebsd-scsi-outgoing; Mon, 7 Oct 1996 09:09:09 -0700 (PDT) Received: from hp.com (hp.com [15.255.152.4]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id JAA08335; Mon, 7 Oct 1996 09:07:44 -0700 (PDT) Received: from srmail.sr.hp.com by hp.com with ESMTP (1.37.109.16/15.5+ECS 3.3) id AA136814452; Mon, 7 Oct 1996 09:07:32 -0700 Received: from hpnmhjw.sr.hp.com by srmail.sr.hp.com with ESMTP (1.37.109.16/15.5+ECS 3.3) id AA220544450; Mon, 7 Oct 1996 09:07:31 -0700 Received: from mina.sr.hp.com by hpnmhjw.sr.hp.com with SMTP (1.37.109.16/15.5+ECS 3.3) id AA002654450; Mon, 7 Oct 1996 09:07:30 -0700 Message-Id: <199610071607.AA002654450@hpnmhjw.sr.hp.com> To: "Justin T. Gibbs" Cc: Brett Glass , BRETT_GLASS@infoworld.com, scsi@freefall.freebsd.org Subject: Re: Problems with Adaptec EISA twin SCSI? Reply-To: darrylo@sr.hp.com In-Reply-To: Your message of "Mon, 07 Oct 1996 06:55:22 PDT." <199610071355.GAA27757@freefall.freebsd.org> Date: Mon, 07 Oct 1996 09:07:29 -0700 From: Darryl Okahata Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > >Only because some of the docs (I forget whether it was in one of the FAQs > >or in the Handbook) said so. They said there was currently no good way > >of doing an automatic upgrade, and that the manual process was painful > >and could cause you to lose things. If there's a way around it, > >I just need to know what it is.... > > Its called, "cd /usr/src; make world". Since you are already running > 2.1.5R, this will work just fine for you. Nothing has changed in /etc > since the release of 2.1.5R on the "stable" branch. I think the FAQ and/or Handbook are referring to upgrading FreeBSD via the precompiled binaries (and not via "make world"). With this method, sysinstall does greatly munge /etc, and I believe this is where the FAQ/handbook says that the manual process is "painful". You certainly could lose things if you are not careful. Basically, /etc gets backed up, and then all of the major files in /etc get blown away and replaced with "new" ones. You then have to go to the backup /etc and hand-merge your changes into the new /etc. Not pretty, but it works. -- Darryl ("Been there, done that") Okahata Internet: darrylo@sr.hp.com DISCLAIMER: this message is the author's personal opinion and does not constitute the support, opinion, or policy of Hewlett-Packard, or of the little green men that have been following him all day. From owner-freebsd-scsi Mon Oct 7 23:02:09 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id XAA17510 for freebsd-scsi-outgoing; Mon, 7 Oct 1996 23:02:09 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (wck-ca11-04.ix.netcom.com [204.31.231.164]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id XAA17501 for ; Mon, 7 Oct 1996 23:02:05 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.7.6/8.6.9) id WAA01013; Mon, 7 Oct 1996 22:57:39 -0700 (PDT) Date: Mon, 7 Oct 1996 22:57:39 -0700 (PDT) Message-Id: <199610080557.WAA01013@silvia.HIP.Berkeley.EDU> To: pjchilds@imforei.apana.org.au CC: bsdscsi@shadows.aeon.net, freebsd-scsi@freebsd.org, ccd@stampede.cs.berkeley.edu In-reply-to: <199609281959.FAA23341@al.imforei.apana.org.au> (message from Peter Childs on Sun, 29 Sep 1996 05:29:13 +0930 (CST)) Subject: Re: striping/mirroring? From: asami@freebsd.org (Satoshi Asami) Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk * I'm really quite interested in how you propose to have a mirrored * system where drives can run at different steeds. Mirror works like * this * * DATA ----> (going to disk) -------+--(and another copy)----+ * | | * disk 1 disk 2 * * So if i send some data to the (mirrored) disk, and disk1 * is significantly faster than disk2, then disk1 will finish the write * first.... so then do i just let it write more? Can you see a * problem with this? What happens when disk1 is miles ahead.. who * is supposed to keep track of all this data thats queued up for * disk2? The OS? And if the disk 1 suddenly fails then you're * screwed, and you fault tolerant system wan't very. In short, the question is whether return from the write system call after writing to only one disk or wait for both disks to complete, right? FreeBSD's ccd uses the latter method, and NetBSD uses the former (at least it was that way when they first took our mirror code). My reasoning was exactly as you mentioned. Satoshi From owner-freebsd-scsi Mon Oct 7 23:03:51 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id XAA17730 for freebsd-scsi-outgoing; Mon, 7 Oct 1996 23:03:51 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (wck-ca11-04.ix.netcom.com [204.31.231.164]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id XAA17685 for ; Mon, 7 Oct 1996 23:03:26 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.7.6/8.6.9) id XAA01024; Mon, 7 Oct 1996 23:01:22 -0700 (PDT) Date: Mon, 7 Oct 1996 23:01:22 -0700 (PDT) Message-Id: <199610080601.XAA01024@silvia.HIP.Berkeley.EDU> To: pjchilds@imforei.apana.org.au CC: bsdscsi@shadows.aeon.net, freebsd-scsi@freebsd.org In-reply-to: <199609290943.TAA22973@al.imforei.apana.org.au> (message from Peter Childs on Sun, 29 Sep 1996 19:13:18 +0930 (CST)) Subject: Re: striping/mirroring? From: asami@freebsd.org (Satoshi Asami) Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk * From: Peter Childs * A fairly generic example of this would be a machine with a dual * SCSI bus system, a small system drive for root and /usr, and then * you data spread over say 4 2gb disks. * * So you would have * * SCSI BUS 1 SCSI BUS 2 * | | * +--> disk 1 +--> disk 3 * | | * +--> disk 2 +--> disk 4 * * Disk 1 and 2 would form a large interleaved data disks, which would * then be mirrored onto disk's 3 and 4. Actually, it would be better if you write your ccd.conf entry as ccd0 CCDF_MIRROR disk1 disk3 disk2 disk4 That would make disk1 and disk3 the data disks and disk2 and disk4 the mirrors. Reads only come from the first half, so we want to spread those between both controllers. * Advantages are reasonable speed, and a measure of safety. If one disk * fails then you can just "turn-off" ccd and continue on with the other * good pair of disks. Yes...but I would recommend you keep at least one spare disk that you can substitute as soon as you find out you have a bad disk. (Ideally, this spare disk should be already connected so you can immediately start your recovery procedure.) Use "dd" to make a copy of the whole disk. Satoshi From owner-freebsd-scsi Mon Oct 7 23:08:00 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id XAA18351 for freebsd-scsi-outgoing; Mon, 7 Oct 1996 23:08:00 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (wck-ca11-04.ix.netcom.com [204.31.231.164]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id XAA18301 for ; Mon, 7 Oct 1996 23:07:34 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.7.6/8.6.9) id XAA01033; Mon, 7 Oct 1996 23:04:07 -0700 (PDT) Date: Mon, 7 Oct 1996 23:04:07 -0700 (PDT) Message-Id: <199610080604.XAA01033@silvia.HIP.Berkeley.EDU> To: bsdscsi@shadows.aeon.net CC: freebsd-scsi@FreeBSD.ORG In-reply-to: <199609270938.MAA20281@shadows.aeon.net> (message from mika ruohotie on Fri, 27 Sep 1996 12:38:32 +0300 (EET DST)) Subject: Re: striping/mirroring? From: asami@FreeBSD.ORG (Satoshi Asami) Sender: owner-freebsd-scsi@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk * From: mika ruohotie * > If you want to maximize sequential access, you'll need about 6 or 7 * > disks to stripe across them, you'll get to about 28 MB/s with the * > option "I586_FAST_BCOPY". For random access, you'll need more like 30 * > drives through the filesystem to max out the motherboard. * * ok * * only 28 mb/s? isnt it possible to get more speed? That's because there is one memory-memory copy that we can't avoid as long as we go through the filesystem. (Since the memory copy speed is 80MB/s and DMA I/O seems to run at around 60MB/s, we should theoretically be able to get something like 1/(1/80 + 1/60) =~ 35 MB/s but I guess there are some overhead too.) Satoshi From owner-freebsd-scsi Tue Oct 8 00:55:13 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id AAA05569 for freebsd-scsi-outgoing; Tue, 8 Oct 1996 00:55:13 -0700 (PDT) Received: from shadows.aeon.net (bsdscsi@shadows.aeon.net [194.100.41.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id AAA05555; Tue, 8 Oct 1996 00:55:06 -0700 (PDT) Received: (from bsdscsi@localhost) by shadows.aeon.net (8.7.5/8.6.9) id KAA03301; Tue, 8 Oct 1996 10:53:57 +0200 (EET) From: mika ruohotie Message-Id: <199610080853.KAA03301@shadows.aeon.net> Subject: Re: striping/mirroring? To: asami@freebsd.org (Satoshi Asami) Date: Tue, 8 Oct 1996 10:53:57 +0200 (EET) Cc: pjchilds@imforei.apana.org.au, freebsd-scsi@freebsd.org In-Reply-To: <199610080601.XAA01024@silvia.HIP.Berkeley.EDU> from Satoshi Asami at "Oct 7, 96 11:01:22 pm" X-Mailer: ELM [version 2.4ME+ PL22 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > > Actually, it would be better if you write your ccd.conf entry as > ccd0 CCDF_MIRROR disk1 disk3 disk2 disk4 > > That would make disk1 and disk3 the data disks and disk2 and disk4 the > mirrors. Reads only come from the first half, so we want to spread > those between both controllers. does this work already? (i cant experiment yet, disks are on purchase list i have to get signed before i get the disks) and am i right assuming (some old post listed these) that 128 is the interleave value that gives me most speed for reads? question, do i have to mirror as many drives as i am striping? since i'd rather strip to 3 and mirror to one... (living close to the edge... =) ) > * Advantages are reasonable speed, and a measure of safety. If one disk > * fails then you can just "turn-off" ccd and continue on with the other > * good pair of disks. it's this simple? nothing lost? sounds cool... =) > Yes...but I would recommend you keep at least one spare disk that you > can substitute as soon as you find out you have a bad disk. (Ideally, > this spare disk should be already connected so you can immediately > start your recovery procedure.) Use "dd" to make a copy of the whole > disk. but if the spare is online, and running there's the chance it is broken already... and since it just hangs there i would not know it until... > Satoshi mickey From owner-freebsd-scsi Wed Oct 9 03:06:08 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id DAA11310 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 03:06:08 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (wck-ca7-17.ix.netcom.com [204.31.231.49]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id DAA11303 for ; Wed, 9 Oct 1996 03:05:59 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.7.6/8.6.9) id DAA04113; Wed, 9 Oct 1996 03:03:34 -0700 (PDT) Date: Wed, 9 Oct 1996 03:03:34 -0700 (PDT) Message-Id: <199610091003.DAA04113@silvia.HIP.Berkeley.EDU> To: bsdscsi@shadows.aeon.net CC: pjchilds@imforei.apana.org.au, freebsd-scsi@freebsd.org In-reply-to: <199610080853.KAA03301@shadows.aeon.net> (message from mika ruohotie on Tue, 8 Oct 1996 10:53:57 +0200 (EET)) Subject: Re: striping/mirroring? From: asami@freebsd.org (Satoshi Asami) Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk * does this work already? (i cant experiment yet, disks are on purchase * list i have to get signed before i get the disks) You mean mirroring? Yes it's working, has been for a few months. * and am i right assuming (some old post listed these) that 128 is the * interleave value that gives me most speed for reads? Depends on what kind of reads you are talking about. For large sequential reads (here the size of individual read()s don't matter -- the "large" refers to the total size that's read in succession) and many disks, something a little smaller is usually better, e.g., 32 or 64. For random reads, it should probably be the size of the read, i.e., if your reads are 16K, then 32 would do the best. (Unless the read sizes are very large, say 1M or so...in which case, treat this as "sequential" as described above.) Of course, I'm assuming the reads occur at offsets that are integer multiples of the read sizes -- if not, I guess you want something a little larger to reduce the chance of a single read falling between two disks. However, since 128 does reasonably well for both reads and writes, that is the size I would recommend for normal (read/write) workloads. * question, do i have to mirror as many drives as i am striping? since i'd * rather strip to 3 and mirror to one... (living close to the edge... =) ) That's not how mirroring works. (Although it is possible to have a 3-disk stripe and one disk for mirroring using a different implementation than our ccd, that won't protect all 3 disks.) You are probably talking about write-ahead logging, which we don't have. When LFS comes back to earth, you may be able to combine ccd and LFS to do something of that sort. * > * Advantages are reasonable speed, and a measure of safety. If one disk * > * fails then you can just "turn-off" ccd and continue on with the other * > * good pair of disks. * * it's this simple? nothing lost? sounds cool... =) That's why you are paying a very high cost on redundancy (50% of diskspace just for mirroring). Of course, it's still cheaper than hardware RAID boxes. :> * but if the spare is online, and running there's the chance it is broken * already... and since it just hangs there i would not know it until... You can use a SCSI command to tell it to stop spinning while it's sitting there as a spare. In which case, I would put the chance of it breaking roughly equal to a disk just waiting in the shipping box. :) Satoshi From owner-freebsd-scsi Wed Oct 9 08:56:55 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id IAA04445 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 08:56:55 -0700 (PDT) Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id IAA04429; Wed, 9 Oct 1996 08:56:48 -0700 (PDT) Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.7.5/8.7.3) id IAA06884; Wed, 9 Oct 1996 08:56:30 -0700 (PDT) From: "Rodney W. Grimes" Message-Id: <199610091556.IAA06884@GndRsh.aac.dev.com> Subject: Re: striping/mirroring? In-Reply-To: <199610091003.DAA04113@silvia.HIP.Berkeley.EDU> from Satoshi Asami at "Oct 9, 96 03:03:34 am" To: asami@FreeBSD.org (Satoshi Asami) Date: Wed, 9 Oct 1996 08:56:30 -0700 (PDT) Cc: bsdscsi@shadows.aeon.net, pjchilds@imforei.apana.org.au, freebsd-scsi@FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk > * does this work already? (i cant experiment yet, disks are on purchase > * list i have to get signed before i get the disks) > > You mean mirroring? Yes it's working, has been for a few months. > > * and am i right assuming (some old post listed these) that 128 is the > * interleave value that gives me most speed for reads? > > Depends on what kind of reads you are talking about. For large > sequential reads (here the size of individual read()s don't matter -- > the "large" refers to the total size that's read in succession) and > many disks, something a little smaller is usually better, e.g., 32 or > 64. > > For random reads, it should probably be the size of the read, i.e., if > your reads are 16K, then 32 would do the best. (Unless the read sizes > are very large, say 1M or so...in which case, treat this as > "sequential" as described above.) Of course, I'm assuming the reads > occur at offsets that are integer multiples of the read sizes -- if > not, I guess you want something a little larger to reduce the chance > of a single read falling between two disks. > > However, since 128 does reasonably well for both reads and writes, > that is the size I would recommend for normal (read/write) workloads. Could you please start recommending CG size interleaves (65536 or there abouts) for people using this for news spools. I have had several clients contact me about abizmal performance and they where using 16 to 128 block interleaves :-( :-( :-(. -- Rod Grimes rgrimes@gndrsh.aac.dev.com From owner-freebsd-scsi Wed Oct 9 09:45:42 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id JAA08345 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 09:45:42 -0700 (PDT) Received: from mail.vividnet.com (mail.vividnet.com [206.149.144.3]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id JAA08337 for ; Wed, 9 Oct 1996 09:45:39 -0700 (PDT) Received: from cancer.vividnet.com (postmaster@mail.vividnet.com) by mail.vividnet.com (8.7.6/8.7.6) with SMTP id JAA04996 for ; Wed, 9 Oct 1996 09:09:32 -0700 (PDT) Date: Wed, 9 Oct 1996 10:00:53 -0700 (PDT) From: Brian Wang To: freebsd-scsi@freebsd.org Subject: Need some advise on (ncr dead ?) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I need some advise from the logs attached at the end of the e-mail. Something definitely is going berserk on this machine's SCSI subsystem. I'm getting bad blocks on a brand new HD, and error messages from the rest. There's a tape drive attach to it, and I'm unable to perform backups (it fails). I'm thinking that it is a bad SCSI card (replacement on its way), what do you scsi experts think? By the way, is there any way to fix "Bad file descriptor" errors on the FS other than restore from a backup set? Server profile: 2.1.5 FreeBSD OS sd0 2 gig IBM DORS sd1 1 gig Quantum Fireball st0 HP 4 gig Dat ASUS Triton I MB P75 CPU Thank you! Sincerely, Brian --logs-- sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0: restart (ncr dead ?). ncr0: restart (ncr dead ?). sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 sd0(ncr0:1:0): Power on, reset, or bus device reset occurred , retries:2 st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. bad block 1516661186, ino 517 uid 0 on /: bad block bad block 249695476, ino 517 uid 0 on /: bad block sd1(ncr0:3:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. sd1(ncr0:3:0): UNIT ATTENTION asc:29,0 sd1(ncr0:3:0): Power on, reset, or bus device reset occurred , retries:4 From owner-freebsd-scsi Wed Oct 9 10:30:29 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id KAA12251 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 10:30:29 -0700 (PDT) Received: from Octopussy (Octopussy.MI.Uni-Koeln.DE [134.95.212.20]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id KAA12235 for ; Wed, 9 Oct 1996 10:30:13 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr3-14.slip.Uni-Koeln.DE) by Octopussy with SMTP id AA22753 (5.67b/IDA-1.5 for ); Wed, 9 Oct 1996 19:29:59 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.7.6/8.6.9) id TAA03464; Wed, 9 Oct 1996 19:29:54 +0200 (MET DST) Message-Id: <199610091729.TAA03464@x14.mi.uni-koeln.de> Date: Wed, 9 Oct 1996 19:29:54 +0200 From: se@zpr.uni-koeln.de (Stefan Esser) To: brian@vividnet.com (Brian Wang) Cc: freebsd-scsi@freebsd.org Subject: Re: Need some advise on (ncr dead ?) In-Reply-To: ; from Brian Wang on Oct 9, 1996 10:00:53 -0700 References: X-Mailer: Mutt 0.45 Mime-Version: 1.0 Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Brian Wang writes: > Hi, > > I need some advise from the logs attached at the end of the > e-mail. Something definitely is going berserk on this machine's SCSI > subsystem. I'm getting bad blocks on a brand new HD, and error messages > from the rest. There's a tape drive attach to it, and I'm unable to > perform backups (it fails). I'm thinking that it is a bad SCSI card > (replacement on its way), what do you scsi experts think? By the way, is Hmm, no, this does not look like a controller problem. But you don't give any information when this failure happens: During the SCSI boot probe, or after the system has been running for some time. (I assume the former ...) You may want to try the boot floppy from the latest SNAP, which was built with slightly different default parameters for certain driver options. I just received my Ultra-Wide DORS 32160T, which will help me to test the Ultra-Wide NCR driver that is expected to become available soon :) (My Tekram DC390F (Ultra-Wide, 53c875 based) host adapter is still on its way, so I could not start the testing, yet.) > --logs-- > sd0(ncr0:1:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. > sd0(ncr0:1:0): UNIT ATTENTION asc:29,0 > sd0(ncr0:1:0): Power on, reset, or bus device reset occurred > , retries:2 > st0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. These are only informational messages. The devices react on the SCSI bus reset they noticed ... > ncr0: restart (ncr dead ?). This is a message that gets printed when no progress is made on some command for more than 10 seconds. > ncr0: restart (ncr dead ?). If there still is no progress, a SCSI bus reset is issued to recover from the apparently locked state. I'll remove this SCSI reset with the next set of patches to the NCR driver, since it does not appear to do much good. You may want to try the boot floppy from the latest SNAP. It contains a slightly different version of the NCR code, as came with 2.1.5, and it may well make a difference. (No, I don't want you to install the SNAP, but if I know it works with that driver, it will be easier to get your system run with 2.1.5.) Regards, STefan From owner-freebsd-scsi Wed Oct 9 12:15:18 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id MAA22890 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 12:15:18 -0700 (PDT) Received: from Octopussy (Octopussy.MI.Uni-Koeln.DE [134.95.212.20]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id MAA22882 for ; Wed, 9 Oct 1996 12:15:15 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr2-43.slip.Uni-Koeln.DE) by Octopussy with SMTP id AA24851 (5.67b/IDA-1.5 for ); Wed, 9 Oct 1996 21:14:43 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.7.6/8.6.9) id VAA03673; Wed, 9 Oct 1996 21:14:38 +0200 (MET DST) Date: Wed, 9 Oct 1996 21:14:38 +0200 (MET DST) Message-Id: <199610091914.VAA03673@x14.mi.uni-koeln.de> From: Stefan Esser To: Richard Tobin Cc: freebsd-scsi@freebsd.org Subject: Re: Cryptic NCR error In-Reply-To: <199610051346.OAA11683@deacon.cogsci.ed.ac.uk> References: <199610051346.OAA11683@deacon.cogsci.ed.ac.uk> Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Richard Tobin writes: > Can anyone decode this error from my NCR 810 SCSI controller? I've > had it several times since I (a) installed a T4000s tape drive (b) > substantially increased the length of my internal SCSI cable (to > 1.5m). Target 1 (I assume that's what ncr0:1 means) is my disk. > > ncr0:1: ERROR (80:14) (0-a7-80) (8/13) @ (c8c:50000000). > script cmd = 740a8700 > reg: da 10 80 13 47 08 01 1f 35 00 81 a7 80 00 07 02. There are several possibilities what went wrong. If you got further such errro messages in some log file, then please send them to my address (and please include a few more lines above and below, if possible). Which version of FreeBSD (or better yet: of "/sys/pci/ncr.c" :) is this ? What kind of device is that T4000 tape ??? Regards, STefan From owner-freebsd-scsi Wed Oct 9 12:24:56 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id MAA23815 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 12:24:56 -0700 (PDT) Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id MAA23805 for ; Wed, 9 Oct 1996 12:24:53 -0700 (PDT) Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.7.5/8.7.3) id MAA07185; Wed, 9 Oct 1996 12:24:30 -0700 (PDT) From: "Rodney W. Grimes" Message-Id: <199610091924.MAA07185@GndRsh.aac.dev.com> Subject: Re: Need some advise on (ncr dead ?) In-Reply-To: from Brian Wang at "Oct 9, 96 10:00:53 am" To: brian@vividnet.com (Brian Wang) Date: Wed, 9 Oct 1996 12:24:30 -0700 (PDT) Cc: freebsd-scsi@freebsd.org X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > Hi, > > I need some advise from the logs attached at the end of the > e-mail. Something definitely is going berserk on this machine's SCSI > subsystem. I'm getting bad blocks on a brand new HD, and error messages > from the rest. There's a tape drive attach to it, and I'm unable to > perform backups (it fails). I'm thinking that it is a bad SCSI card > (replacement on its way), what do you scsi experts think? By the way, is > there any way to fix "Bad file descriptor" errors on the FS other than > restore from a backup set? > > Server profile: > 2.1.5 FreeBSD OS > sd0 2 gig IBM DORS Having recently had someone bring me 3 of these to build systems with that all turned up seriously damaged I suspect you have a bad drive here. Try to run a low level veryify on it, oopsss... ncr doesn't come with a ``verify'' tool, well, do a low level format using SCSIFMT.EXE. I suspect you'll find a drive with 1000's of bad blocks on it and that likes to spin down repeadly. NOTE TO ALL OTHERS:: BE WARRY, there has been a product dump of IBM 2G DORS drives into the grey market, the ones I looked at had been physically damaged (broken scsi connectors) and had 100's if not 1000's of bad sectors on them. These drives are going at prices that look to good to be true ($350 and below), and it IS TOO GOOD TO BE TRUE. [Messages deleted, but that was the first messages I got before I started to investigate the integritty of the drive itself, I could here the drive spin down just before a burst of errors occured :-(] -- Rod Grimes rgrimes@gndrsh.aac.dev.com From owner-freebsd-scsi Wed Oct 9 13:13:30 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id NAA28077 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 13:13:30 -0700 (PDT) Received: from mail.vividnet.com (mail.vividnet.com [206.149.144.3]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id NAA28066 for ; Wed, 9 Oct 1996 13:13:26 -0700 (PDT) Received: from cancer.vividnet.com (postmaster@mail.vividnet.com) by mail.vividnet.com (8.7.6/8.7.6) with SMTP id MAA07403; Wed, 9 Oct 1996 12:36:59 -0700 (PDT) Date: Wed, 9 Oct 1996 13:28:29 -0700 (PDT) From: Brian Wang Reply-To: Brian Wang To: Stefan Esser cc: freebsd-scsi@freebsd.org Subject: Re: Need some advise on (ncr dead ?) In-Reply-To: <199610091729.TAA03464@x14.mi.uni-koeln.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Wed, 9 Oct 1996, Stefan Esser wrote: > Brian Wang writes: > > Hi, > > > > I need some advise from the logs attached at the end of the > > e-mail. Something definitely is going berserk on this machine's SCSI > > subsystem. I'm getting bad blocks on a brand new HD, and error messages > > from the rest. There's a tape drive attach to it, and I'm unable to > > perform backups (it fails). I'm thinking that it is a bad SCSI card > > (replacement on its way), what do you scsi experts think? By the way, is > > Hmm, no, this does not look like a controller problem. > But you don't give any information when this failure > happens: During the SCSI boot probe, or after the system > has been running for some time. (I assume the former ...) 12:19PM up 27 days, 2:55, 1 user, load averages: 0.08, 0.05, 0.01 It's up, and running fine until this week. I haven't reboot it yet, and won't do it until I have everything ready (backups/parts). This is our pop/mail server :( On another reply to my plead for help, Rodney W. Grimes also suggested that it could be that damn IBM DORS HD. My best bet is to replace the HD. Thank you all! Sincerely, Brian From owner-freebsd-scsi Wed Oct 9 13:18:06 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id NAA28579 for freebsd-scsi-outgoing; Wed, 9 Oct 1996 13:18:06 -0700 (PDT) Received: from mail.vividnet.com (mail.vividnet.com [206.149.144.3]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id NAA28571 for ; Wed, 9 Oct 1996 13:18:01 -0700 (PDT) Received: from cancer.vividnet.com (postmaster@mail.vividnet.com) by mail.vividnet.com (8.7.6/8.7.6) with SMTP id MAA07464; Wed, 9 Oct 1996 12:41:49 -0700 (PDT) Date: Wed, 9 Oct 1996 13:33:19 -0700 (PDT) From: Brian Wang To: "Rodney W. Grimes" cc: freebsd-scsi@freebsd.org Subject: Re: Need some advise on (ncr dead ?) In-Reply-To: <199610091924.MAA07185@GndRsh.aac.dev.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Wed, 9 Oct 1996, Rodney W. Grimes wrote: > Having recently had someone bring me 3 of these to build systems with > that all turned up seriously damaged I suspect you have a bad drive here. > > Try to run a low level veryify on it, oopsss... ncr doesn't come > with a ``verify'' tool, well, do a low level format using SCSIFMT.EXE. > > I suspect you'll find a drive with 1000's of bad blocks on it and > that likes to spin down repeadly. > > > NOTE TO ALL OTHERS:: BE WARRY, there has been a product dump of IBM > 2G DORS drives into the grey market, the ones I looked at had been > physically damaged (broken scsi connectors) and had 100's if not > 1000's of bad sectors on them. These drives are going at prices that > look to good to be true ($350 and below), and it IS TOO GOOD TO BE TRUE. > -- > Rod Grimes rgrimes@gndrsh.aac.dev.com > Thanks for the info! I'm going to replace that damn HD. Sincerely, Brian From owner-freebsd-scsi Thu Oct 10 10:34:14 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id KAA04741 for freebsd-scsi-outgoing; Thu, 10 Oct 1996 10:34:14 -0700 (PDT) Received: from agora.rdrop.com (root@agora.rdrop.com [199.2.210.241]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id KAA04734 for ; Thu, 10 Oct 1996 10:34:12 -0700 (PDT) Received: from unicorn.uk1.vbc.net by agora.rdrop.com with smtp (Smail3.1.29.1 #17) id m0vBOzl-00094ZC; Thu, 10 Oct 96 10:34 PDT Received: (from gordon@localhost) by unicorn.uk1.vbc.net (8.7.3/8.7.3) id SAA06248; Thu, 10 Oct 1996 18:08:32 +0100 Date: Thu, 10 Oct 1996 18:08:30 +0100 (BST) From: Gordon Henderson X-Sender: gordon@unicorn To: freebsd-scsi@freebsd.org Subject: Buslogic controller, Sync mode & a SCSI disk error Message-ID: Distribution: world Organization: Home for lost Drogons MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I have a Bizarre set of problems... Heres the setup: ASUS P120, 128MB RAM, 2 Buslogic 946C controllers each with 3 identical 2GB drives. Running FreeBSD 2.1.5R. Firstly: Boot messages which I find rather odd: [I've cut some of the verbage] /kernel: bt0 rev 0 int a irq 10 on pci0:11 /kernel: bt0: Bt946C/ 0-(32bit) bus /kernel: bt0: reading board settings, busmastering, int=10 /kernel: bt0: version 4.25J, fast sync, parity, 32 mbxs, 3 2 ccbs /kernel: bt0: targ 0 sync rate=10.00MB/s(100ns), offset=15 /kernel: bt0: targ 1 sync rate=10.00MB/s(100ns), offset=15 /kernel: bt0: targ 2 sync rate=10.00MB/s(100ns), offset=15 /kernel: bt0: Using Strict Round robin scheme /kernel: bt0 waiting for scsi devices to settle [device probing snipped] /kernel: bt1 rev 0 int a irq 11 on pci0:12 /kernel: bt1: Bt946C/ 0-(32bit) bus /kernel: bt1: reading board settings, busmastering, int=11 /kernel: bt1: version 4.28D, async only, parity, 32 mbxs, 32 ccbs /kernel: bt1: targ 0 async /kernel: bt1: targ 1 async /kernel: bt1: targ 2 async /kernel: bt1: Using Strict Round robin scheme /kernel: bt1 waiting for scsi devices to settle So - 2 different versions of the Buslogic board, and the most recent one doesn't come up in Sync mode.... Any reason why? (According to Buslogic, that version of board firmware is good and I should use Linux instead of FreeBSD - I did boot Linux on it once while testing it and Linux correctly enabled all devices in 10MB/sec sync mode - Why doesn't FreeBSD?) 2nd problem: One of the disks seems to have a fault. Heres the errors from the messages file: /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 Unrecovered read error /kernel: , retries:4 /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 Unrecovered read error /kernel: , retries:3 /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 Unrecovered read error /kernel: , retries:2 /kernel: bt0: Try to abort /kernel: bt0: not taking commands! /kernel: Debugger("bt742a") called. /kernel: bt0: Abort Operation has timed out at this stage the machine rebooted it's self, fsck'd ok and carried on. (It's a news server). That disk isn't used for swap so it was a file read that caused the failure. Why should reading a duff sector cause the system to crash? So, anyone any ideas how to fix the error, why the bt driver crashes, and why my 2nd controller doesn't come up in sync mode? And exactly what are you supposed to do on a SCSI error anyway? How do I mark the block bad, I know about the scsi command - Used: scsi -f /dev/sd1 -m 1 and got: (amongst other things): AWRE (Auto Write Reallocation Enbld): 1 ARRE (Auto Read Reallocation Enbld): 1 So the disk controller is supposed to automagically re-allocate bad blocks - if thats the case, why am I seeing faults and why is it crashing the machine? If that isn't the case, how do I scan the disk for bad blocks and mark them bad. I've been told that bad144 is only for IDE drives, is that right? If so, whats for SCSI drives? Any help would be gratefully appreciated - I'm sure I'm not the only one in the world with a duff sector or 2 on a FreeBSD SCSI drive! Gordon From owner-freebsd-scsi Thu Oct 10 12:20:43 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id MAA10047 for freebsd-scsi-outgoing; Thu, 10 Oct 1996 12:20:43 -0700 (PDT) Received: from hda.com (ip51-max1-fitch.zipnet.net [199.232.245.51]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id MAA10042 for ; Thu, 10 Oct 1996 12:20:36 -0700 (PDT) Received: (from dufault@localhost) by hda.com (8.6.11/8.6.9) id PAA11730; Tue, 8 Oct 1996 15:28:59 -0400 From: Peter Dufault Message-Id: <199610081928.PAA11730@hda.com> Subject: Re: Buslogic controller, Sync mode & a SCSI disk error To: gordon@drogon.net (Gordon Henderson) Date: Tue, 8 Oct 1996 15:28:58 -0400 (EDT) Cc: freebsd-scsi@freebsd.org In-Reply-To: from "Gordon Henderson" at Oct 10, 96 06:08:30 pm Reply-to: hdalog@zipnet.net X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > So - 2 different versions of the Buslogic board, and the most recent one > doesn't come up in Sync mode.... Any reason why? (According to Buslogic, > that version of board firmware is good and I should use Linux instead of > FreeBSD - I did boot Linux on it once while testing it and Linux correctly > enabled all devices in 10MB/sec sync mode - Why doesn't FreeBSD?) In bt.c the driver is reading back the board setup and interpreting the setting as being for async only. Either the Linux driver ignores the setting or the FreeBSD driver misinterprets the setting. Do you have a Buslogic utility to set up the board? See if it has a setting to permit synchronous negotiation. > > 2nd problem: One of the disks seems to have a fault. Heres the errors > from the messages file: > > /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 > Unrecovered read error > /kernel: , retries:4 > /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 > Unrecovered read error > /kernel: , retries:3 > /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 > Unrecovered read error > /kernel: , retries:2 > /kernel: bt0: Try to abort > /kernel: bt0: not taking commands! > /kernel: Debugger("bt742a") called. > /kernel: bt0: Abort Operation has timed out > > at this stage the machine rebooted it's self, fsck'd ok and carried on. > (It's a news server). That disk isn't used for swap so it was a file read > that caused the failure. Why should reading a duff sector cause the system > to crash? A SCSI disk won't slip the sector on an unrecovered read error - it doesn't know what to put there. You'll have to write something there, preferably by restoring the partition, or by using "dd" to write something to that sector once you've decided what to write there. If you do decide to restore the partition blast it full of zeros using dd first to make sure any failing sectors get mapped out on the write failure. The problem with bt0 locking up and rebooting on this error is a bug. -- Peter Dufault Real-Time Machine Control and Simulation HD Associates, Inc. Voice: 508 433 6936 dufault@hda.com Fax: 508 433 5267 From owner-freebsd-scsi Thu Oct 10 14:51:34 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id OAA18958 for freebsd-scsi-outgoing; Thu, 10 Oct 1996 14:51:34 -0700 (PDT) Received: from GndRsh.aac.dev.com ([198.145.92.241]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id OAA18949 for ; Thu, 10 Oct 1996 14:51:24 -0700 (PDT) Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.7.5/8.7.3) id OAA09130; Thu, 10 Oct 1996 14:50:52 -0700 (PDT) From: "Rodney W. Grimes" Message-Id: <199610102150.OAA09130@GndRsh.aac.dev.com> Subject: Re: Buslogic controller, Sync mode & a SCSI disk error In-Reply-To: from Gordon Henderson at "Oct 10, 96 06:08:30 pm" To: gordon@drogon.net (Gordon Henderson) Date: Thu, 10 Oct 1996 14:50:51 -0700 (PDT) Cc: freebsd-scsi@freebsd.org X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > > I have a Bizarre set of problems... > > Heres the setup: ASUS P120, 128MB RAM, 2 Buslogic 946C controllers each > with 3 identical 2GB drives. Running FreeBSD 2.1.5R. > > Firstly: Boot messages which I find rather odd: > > [I've cut some of the verbage] [I cut more of it :-)] > /kernel: bt0: version 4.25J, fast sync, parity, 32 mbxs, 3 2 ccbs > /kernel: bt1: version 4.28D, async only, parity, 32 mbxs, 32 ccbs > > So - 2 different versions of the Buslogic board, and the most recent one > doesn't come up in Sync mode.... Any reason why? Yes, Buslogic made some additional bit settings you have to do to get the bt946 into sync mode. > (According to Buslogic, > that version of board firmware is good and I should use Linux instead of > FreeBSD - Would you please give me the persons name at Buslogic who said this, I want to scream very loudly at them for that one. After I wasted 100's of hours trying to get the correct information from them to fix the problem they start pulling this kind of shit has my shirt in an uproar... [Okay, Justin, I'll back off, but I want a name, and if I hear that name again I am going to make some calls!!] > I did boot Linux on it once while testing it and Linux correctly > enabled all devices in 10MB/sec sync mode - Why doesn't FreeBSD?) Because the Linux developer had access to resources that have been denied up until this point in time. Justin Gibbs has supposedly broken down that road and the information needed to fix this and a pile of other bugs should be forth coming. > 2nd problem: One of the disks seems to have a fault. Heres the errors > from the messages file: > > /kernel: sd1(bt0:1:0): MEDIUM ERROR info:272133 asc:11,0 > Unrecovered read error ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ARRE can _NOT_ correct an UNRECOVERED read error, as that would mean data loss. ARRE and AWRE can only do thier trick if no data would be lost. ... > > So the disk controller is supposed to automagically re-allocate bad blocks > - if thats the case, why am I seeing faults and why is it crashing the > machine? the disk DRIVE is suppose to automagically re-allocate bad blocks, if and only if, no data would be lost. If it can't read the old block at all (trying all methods of data recovery such as retries, head offsetting, etc) it has no choise but to return an error to the OS. You can try 4 or 5 times doing this and see if by chance enough forced retries can read the block a recovery would happen: dd if=/dev/rsdX of=/dev/null bs=32768 BACKUPS ARE STRONGLY RECOMMENDED BEFORE DOING THIS! If that fails you are going to have to run a low level verify operation and tell it to remap the block that it could not read, then run fsck and hope to hell it was just a file system block and not meta data. -- Rod Grimes rgrimes@gndrsh.aac.dev.com From owner-freebsd-scsi Fri Oct 11 07:00:46 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id HAA16831 for freebsd-scsi-outgoing; Fri, 11 Oct 1996 07:00:46 -0700 (PDT) Received: from brasil.moneng.mei.com (brasil.moneng.mei.com [151.186.109.160]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id GAA16678 for ; Fri, 11 Oct 1996 06:59:09 -0700 (PDT) Received: (from jgreco@localhost) by brasil.moneng.mei.com (8.7.Beta.1/8.7.Beta.1) id IAA20953; Fri, 11 Oct 1996 08:57:50 -0500 From: Joe Greco Message-Id: <199610111357.IAA20953@brasil.moneng.mei.com> Subject: Re: Buslogic controller, Sync mode & a SCSI disk error To: rgrimes@GndRsh.aac.dev.com (Rodney W. Grimes) Date: Fri, 11 Oct 1996 08:57:50 -0500 (CDT) Cc: gordon@drogon.net, freebsd-scsi@FREEBSD.ORG In-Reply-To: <199610102150.OAA09130@GndRsh.aac.dev.com> from "Rodney W. Grimes" at Oct 10, 96 02:50:51 pm X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-scsi@FREEBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > BACKUPS ARE STRONGLY RECOMMENDED BEFORE DOING THIS! > If that fails you are going to have to run a low level verify operation > and tell it to remap the block that it could not read, then run fsck > and hope to hell it was just a file system block and not meta data. Hi Rod, Is there a good tool to do this sort of stuff under FreeBSD (besides doing it by hand with the scsi command)? The NCR controllers in use around here do not have a built in BIOS utility (like the Adaptecs do) to do low level formatting and verification/bad block remapping. I've typically found it more convenient to simply disconnect the SCSI bus from the NCR and hook it up to an Adaptec 15x2 and do whatever. Unfortunately that requires physical intervention. Since I am heavily into remote administration, I would like to see a tool that was able to do similar things, running under FreeBSD. However, I do not have the SCSI knowledge to write such a tool, and I have not seen any references to an existing tool already written. Any ideas, comments, etc.? Thanks, ... JG From owner-freebsd-scsi Fri Oct 11 20:47:21 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id UAA21511 for freebsd-scsi-outgoing; Fri, 11 Oct 1996 20:47:21 -0700 (PDT) Received: from frmug.org (frmug-gw.frmug.org [193.56.58.252]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id UAA21504 for ; Fri, 11 Oct 1996 20:47:13 -0700 (PDT) Received: (from uucp@localhost) by frmug.org (8.6.8/8.6.9) with UUCP id FAA22976 for freebsd-scsi@freebsd.org; Sat, 12 Oct 1996 05:47:04 +0200 Received: (from elrond@localhost) by imladris.frmug.fr.net (8.7.5/8.7.3) id IAA02796 for freebsd-scsi@freebsd.org; Sat, 12 Oct 1996 08:46:43 +0200 (MET DST) From: Bertrand Petit Message-Id: <199610120646.IAA02796@imladris.frmug.fr.net> Subject: Success report: NOMAI MCD540 To: freebsd-scsi@freebsd.org Date: Sat, 12 Oct 1996 08:46:42 +0200 (MET DST) X-roberto: ce sont X-remibp: Ah! It is marching. X-Mailer: ELM [version 2.4 PL20] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I sucessfuly hooked a NOMAI MCD540 removable disk unit to a FreeBSD 2.1.0-RELEASE system. I got some troubles to make it work, I got tons of bus resets and NCR crashes likes this: sd5(ncr0:5:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8. ncr0:5: ERROR (80:100) (8-2a-0) (8/13) @ (544:900b0000). script cmd = 910a0000 reg: da 10 80 13 47 08 05 1f 01 08 05 2a 80 00 02 00. ncr0: handshake timeout sd5(ncr0:5:0): COMMAND FAILED (6 ff) @f0c04c00. It was solved by a main board BIOS update on an Asus P/I-55TP4XE with an Asus PCI-SCSI2000 adaptater. The last version of the BIOS available from the Asus ftp site or mirrors, in the file TCX5I020.ZIP. I can't explain why it work with this release and not the the older one as the NCR bios seems to be the same between the two version. Anyone got some ideas about this? -- %!ps | Bertrand Petit alias | elrond@imladris.frmug.fr.net | Cette zone est | % | >elrond le demi-Elfe< | elrond@freenix.fr | a louer | 550 0 translate 90 rotate/NewCenturySchlbk-Bold findfont 690 scalefont setfont -5 15 moveto(42)true charpath clip/Helvetica-Bold findfont 8 scalefont setfont 0 9 800{dup 360 mod 360 div 1 1 sethsbcolor/y exch def rand 10 mod neg 11 800 {y moveto(42)show}for}for clippath 4 setlinewidth 0 setgray stroke showpage From owner-freebsd-scsi Fri Oct 11 21:27:03 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id VAA23719 for freebsd-scsi-outgoing; Fri, 11 Oct 1996 21:27:03 -0700 (PDT) Received: from brasil.moneng.mei.com (brasil.moneng.mei.com [151.186.109.160]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id VAA23711 for ; Fri, 11 Oct 1996 21:26:58 -0700 (PDT) Received: (from jgreco@localhost) by brasil.moneng.mei.com (8.7.Beta.1/8.7.Beta.1) id XAA21911; Fri, 11 Oct 1996 23:25:50 -0500 From: Joe Greco Message-Id: <199610120425.XAA21911@brasil.moneng.mei.com> Subject: Re: Success report: NOMAI MCD540 To: elrond@imladris.frmug.fr.net (Bertrand Petit) Date: Fri, 11 Oct 1996 23:25:49 -0500 (CDT) Cc: freebsd-scsi@freebsd.org In-Reply-To: <199610120646.IAA02796@imladris.frmug.fr.net> from "Bertrand Petit" at Oct 12, 96 08:46:42 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > It was solved by a main board BIOS update on an Asus > P/I-55TP4XE with an Asus PCI-SCSI2000 adaptater. The last version of > the BIOS available from the Asus ftp site or mirrors, in the file > TCX5I020.ZIP. Is this the ASUS PCI-DA2000, or the ASUS PCI-SC200? I would like to hear more about the DA2000 if it is compatible with FreeBSD :-) :-) ... JG From owner-freebsd-scsi Sat Oct 12 22:32:48 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA07155 for freebsd-scsi-outgoing; Sat, 12 Oct 1996 22:32:48 -0700 (PDT) Received: from post.io.org (post.io.org [198.133.36.6]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA07133 for ; Sat, 12 Oct 1996 22:32:30 -0700 (PDT) Received: from zap.io.org (taob@zap.io.org [198.133.36.81]) by post.io.org (8.7.5/8.7.3) with SMTP id BAA11702; Sun, 13 Oct 1996 01:32:23 -0400 (EDT) Date: Sun, 13 Oct 1996 01:32:23 -0400 (EDT) From: Brian Tao To: FREEBSD-SCSI-L cc: jeff@openstore.com Subject: RAID benchmarks: CMD and RAIDION Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@freebsd.org X-Loop: FreeBSD.org Precedence: bulk This is a followup to some RAID benchmark results I had posted three weeks ago. The host system and disk subsystems have changed, so I redid the iozone, bonnie and touch benchmarks. The three contenders are a wide Seagate 5400 rpm 1GB Hawk drive, a CMD-5500 RAID controller with 5 narrow Seagate 7200 rpm 4GB Barracuda drives, and a Streamlogic RAIDION LT RAID controller with 3 narrow Micropolis 4GB drives (I was unable to get specifics on the drives themselves). The CMD-5500 controller has 64MB of RAM, of which about 62MB is usable as write-back cache. The RAIDION has 8MB of RAM, of which about 5.5MB is usable as write-back cache (I mistakenly said it only had 4MB in my previous post). Read-ahead was turned off on both RAIDs. The RAIDs and the Seagate Hawk are hooked up to an Adaptec 2940UW controller on a 200-MHz Pentium Pro host with 128MB. There were no special SCSI tweaks made to the kernel, other than including the ahc driver. >>>>> FreeBSD 2.2-960801-SNAP #0: Tue Sep 24 06:28:45 EDT 1996 root@install1.io.org:/mnt/sys/compile/BEAST Calibrating clock(s) relative to mc146818A clock... i586 clock: 199307170 Hz, i8254 clock: 1193171 Hz CPU: Pentium Pro (199.30-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0x617 Stepping=7 Features=0xf9ff,MTRR,PGE,MCA,CMOV> real memory = 134217728 (131072K bytes) avail memory = 130293760 (127240K bytes) Probing for devices on PCI bus 0: chip0 rev 2 on pci0:0 chip1 rev 1 on pci0:1:0 pci0:1:1: Intel Corporation, device=0x7010, class=storage (ide) [no driver assigned] de0 rev 18 int a irq 10 on pci0:10 de0: SMC 9332 DC21140 [10-100Mb/s] pass 1.2 de0: address 00:00:c0:07:06:e7 de0: enabling 10baseT port ahc0 rev 0 int a irq 11 on pci0:11 ahc0: aic7880 Wide Channel, SCSI Id=7, 16 SCBs ahc0 waiting for scsi devices to settle (ahc0:1:0): "SEAGATE ST31230W 0300" type 0 fixed SCSI 2 sd0(ahc0:1:0): Direct-Access 1010MB (2069860 512 byte sectors) sd0(ahc0:1:0): with 3992 cyls, 5 heads, and an average 103 sectors/track (ahc0:2:0): "OSS INFINITY A4-1" type 0 fixed SCSI 2 sd1(ahc0:2:0): Direct-Access 8191MB (16775168 512 byte sectors) sd1(ahc0:2:0): with 16382 cyls, 16 heads, and an average 64 sectors/track (ahc0:3:0): "MICROP LTX 011000 7t38" type 0 fixed SCSI 2 sd1(ahc0:3:0): Direct-Access 8189MB (16771841 512 byte sectors) sd1(ahc0:3:0): with 8189 cyls, 64 heads, and an average 32 sectors/track <<<<< There was a problem when both RAIDs were daisychained together. The boot manager would only see a single bootable disk (even though both were fdisked as "startable"). Pressing F1 to boot the first disk (presumably the Seagate Hawk) would hang the boot manager for about 30 seconds, then the F1 prompt would reappear. That's why both RAID units appear as sd1 (I only hooked one up at a time). Dunno if anyone has seen that particular problem before... Although the CMD-5500 shipped with five 4GB drives, I only allocated three of them for the comparison tests. Filesystem size is approximately 8GB, RAID 5 (parity staggered across all drives), 64K block size. The RAIDION's maximum block size is 64K, the CMD's is 512K. The benchmark results displayed below are iozone on a 256MB file, bonnie on a 256MB file, and the timed output for the following command line: time touch `jot 10000` ; time touch `jot 10000` ; time rm `jot 10000` I'm working with the RAIDION vendor now to figure out why their product has such a poor showing in the sequential throughput tests. ;-) Turning the RAIDION down to RAID 0, same block size, results in a small increase in performance (~1.3MB/s write, ~6.3MB/s read). This is probably due to having three disks dedicated to reading and writing rather than any overhead save from the calculation of parity. The CMD has a load monitor on its R3000-based CPU. Interestingly, it was doing more work at RAID 0 (~33% idle) than it was at RAID 5 (~65% idle). The on-site tech figures the parity calculation is actually handled by an ASIC, and not the CPU itself. With all five drives in RAID 0, I was able to sustain a 100% load on the CPU, pumping ~14MB/s writing and ~11MB/s reading. CMD claims a peak of 17MB/s (with many more disks and fast/wide disks, I imagine). Both have utilities accessible via a 9600 bps VT100 terminal. The CMD one is far more comprehensive, with performance and environmental monitoring (updated in real time) and lots of status reports on the health of the system. That's about all I have to say right now; the numbers pretty much speak for themselves (especially the file creation and deletion... write-back cache helps *immensely* here). For those interested in buying a RAID product, you are looking at between $750 to $1250 US per gigabyte of storage for a system that offers you protection through parity, online spares, hot swappable drives, etc. Pretty hefty premium, but then compare that to the cost of losing several gigabytes of mail spool... I'd be interested if someone could provide similar benchmarks with a Mylex or a DPT RAID controller. >>>>> BLOCK ---- SINGLE ---- --- RAIDION --- ----- CMD ----- SIZE WRITE READ WRITE READ WRITE READ 512 3518302 5572451 988854 5132149 7573228 8915344 1024 3536771 5746736 992367 5044742 7531726 9344503 2048 3531319 5796177 999846 5084305 7515253 9557646 4096 3511111 5800090 1000895 5069303 7313694 9447274 8192 3570955 5811863 996541 5261022 7543301 9528490 16384 3553597 5592405 997640 5036607 7530076 9554988 32768 3566508 5808916 1000982 5053645 7403520 9447274 65536 3580631 5802049 981398 5100911 7578239 9541721 131072 3516861 5740016 976545 5037346 7454922 9344503 262144 3448734 5249769 971327 5057365 7335554 9263882 -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU single 256 3473 29.0 3555 7.6 1659 5.4 4782 44.6 3716 5.6 105.2 2.0 raidion 256 985 8.2 954 1.9 772 2.2 4754 44.3 5110 5.6 120.8 1.7 cmd 256 7213 59.4 7176 16.1 3521 11.1 7328 68.4 6136 6.7 187.8 2.7 SINGLE touch: 0.277u 56.454s 3:54.02 24.2% 10+170k 166+20314io 14pf+0w retouch: 0.193u 2.796s 1:49.61 2.7% 17+190k 2+10000io 0pf+0w unlink: 0.199u 4.792s 1:52.40 4.4% 167+226k 1+10000io 6pf+0w RAIDION touch: 0.245u 57.470s 1:16.07 75.8% 10+171k 159+20314io 15pf+0w retouch: 0.174u 2.797s 0:11.59 25.5% 16+176k 2+10000io 0pf+0w unlink: 0.171u 4.838s 0:13.55 36.9% 160+216k 1+10000io 3pf+0w CMD touch: 0.192u 56.159s 1:08.75 81.9% 10+169k 166+20314io 29pf+0w retouch: 0.187u 2.764s 0:09.25 31.7% 16+185k 1+10000io 0pf+0w unlink: 0.216u 4.757s 0:11.07 44.8% 164+220k 2+10000io 0pf+0w <<<<< -- Brian Tao (BT300, taob@io.org, taob@ican.net) Senior Systems and Network Administrator, Internet Canada Corp. "Though this be madness, yet there is method in't"