Re: [SAGE] FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
On Wed, Jul 09, 2003 at 08:53:02PM -0500 or thereabouts, Dustin Puryear wrote: Before reading below: I am considering a new install rather than an upgrade of our servers. However, now I just want to beat this problem. :) At 02:19 PM 7/8/2003 +, Phil Pennock wrote: On 2003-07-08 at 14:09 +, Phil Pennock wrote: There was a fairly major update to the IDE disk device handlers which required new device nodes. Bringing in the new MAKEDEV script and running that to create the disk devices will probably let you boot on the new kernel and try to repair things from there. Okay, I'll go get more coffee after sending this. You noticed this one. Sorry. There are two mergemaster steps, one which does the minimum needed to let the rest of the build continue. Yes, I ran 'mergemaster -p' but found that it really only merged my passwd and group files with the new ones. Unless it did some other things behind the scenes then there is no fix here. mergemaster -p -- run before installworld or (preferably) buildworld -- good. The issue is, as you note, the device stuff. As I said though, just pull in MAKEDEV manually and run it. That will fix this. Either sh MAKEDEV all or sh MAKEDEV ad0 ... I keep thinking it's the device stuff myself, but even MAKEDEV didn't do the trick (interesting notes below this stuff): # cd /usr/src # make update # /usr/src/usr.sbin/mergemaster/mergemaster.sh -p # make buildworld # make buildkernel # make installkernel You might have to installworld before rebooting -- I know I always do. # cd /etc # mv MAKEDEV MAKEDEV.old # cp /usr/src/etc/MAKEDEV . # sh MAKEDEV all # ls ad4s1* || sh MAKEDEV ad4s1 # reboot Upon reboot I get my favorite lines: blah, blah blah, blah ar0: ... ATA RAID1 array [...] status: READY subdisks: 0 READY ad4: ... at ata2-master UDMA100 1 READY ad6: ... at ata3-master UDMA100 acd0: CDROM ... at ata0-master PIO4 Mounting root from ufs:/dev/ad4s1a Root mount failed: 16 Manual root filesystem specification: blah, blah According to intro(2), error 16 is EBUSY (Device busy). I think the kernel is not letting you use a device in the RAID. After all, it says subdisks: ad4 ad6 and you're trying to boot from ad4. Could that be it? HTH, -- Josh ...snip... Mounting root from ufs:/dev/ad4s1a --- Dustin Puryear [EMAIL PROTECTED] Puryear Information Technology, LLC http://www.puryear-it.com Providing expertise in the management, integration, and security of Windows and UNIX systems, networks, and applications. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED] ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: [SAGE] FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
Before reading below: I am considering a new install rather than an upgrade of our servers. However, now I just want to beat this problem. :) At 02:19 PM 7/8/2003 +, Phil Pennock wrote: On 2003-07-08 at 14:09 +, Phil Pennock wrote: There was a fairly major update to the IDE disk device handlers which required new device nodes. Bringing in the new MAKEDEV script and running that to create the disk devices will probably let you boot on the new kernel and try to repair things from there. Okay, I'll go get more coffee after sending this. You noticed this one. Sorry. There are two mergemaster steps, one which does the minimum needed to let the rest of the build continue. Yes, I ran 'mergemaster -p' but found that it really only merged my passwd and group files with the new ones. Unless it did some other things behind the scenes then there is no fix here. The issue is, as you note, the device stuff. As I said though, just pull in MAKEDEV manually and run it. That will fix this. Either sh MAKEDEV all or sh MAKEDEV ad0 ... I keep thinking it's the device stuff myself, but even MAKEDEV didn't do the trick (interesting notes below this stuff): # cd /usr/src # make update # /usr/src/usr.sbin/mergemaster/mergemaster.sh -p # make buildworld # make buildkernel # make installkernel # cd /etc # mv MAKEDEV MAKEDEV.old # cp /usr/src/etc/MAKEDEV . # sh MAKEDEV all # reboot Upon reboot I get my favorite lines: blah, blah blah, blah ar0: ... ATA RAID1 array [...] status: READY subdisks: 0 READY ad4: ... at ata2-master UDMA100 1 READY ad6: ... at ata3-master UDMA100 acd0: CDROM ... at ata0-master PIO4 Mounting root from ufs:/dev/ad4s1a Root mount failed: 16 Manual root filesystem specification: blah, blah Okay, so I reboot, but this time I load kernel.good and the system comes up fine with the old 4.4-REL kernel: ar0: ... ATA RAID1 array [...] status: READY subdisks: ad4: ... at ata2-master UDMA100 ad6: ... at ata3-master UDMA100 acd0: CDROM ... at ata0-master using PIO4 Oh, and I found that this server has a HighPoint RAID controller which can do RAID 0, 1, 0+1, and JBOD. Looking in my array configuration I see that we have: ChannelMode Status Primary Master: Mirror (RAID 1) for array #0 UDMA 5HDD0 Secondary Master: Mirror ... UDMA 5Hidden I initially said that this was a Promise controller. Here is my /etc/fstab for /: /dev/ad4s1a / ufs rw 1 1 Following is /var/run/dmesg.log after booting from kernel.good: Copyright (c) 1992-2001 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD 4.4-RELEASE #1: Sun Jun 30 21:23:39 CDT 2002 [EMAIL PROTECTED]:/usr/src/sys/compile/WWW Timecounter i8254 frequency 1193182 Hz Timecounter TSC frequency 1494468860 Hz CPU: Pentium 4 (1494.47-MHz 686-class CPU) Origin = GenuineIntel Id = 0xf07 Stepping = 7 Features=0x3febf9ffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,b28,ACC real memory = 268369920 (262080K bytes) avail memory = 257404928 (251372K bytes) Preloaded elf kernel kernel.good at 0xc0375000. md0: Malloc disk Using $PIR table, 12 entries at 0xc00fdea0 npx0: math processor on motherboard npx0: INT 16 interface pcib0: Host to PCI bridge on motherboard pci0: PCI bus on pcib0 pcib1: PCI to PCI bridge (vendor=8086 device=2532) at device 1.0 on pci0 pci1: PCI bus on pcib1 pci1: NVidia Riva Vanta TNT2 graphics accelerator at 0.0 irq 12 pcib2: Intel 82801BA/BAM (ICH2) Hub to PCI bridge at device 30.0 on pci0 pci2: PCI bus on pcib2 rl0: D-Link DFE-530TX+ 10/100BaseTX port 0x9000-0x90ff mem 0xd900-0xd9ff irq 11 at device 3.0 on pci2 rl0: Ethernet address: 00:50:ba:60:7b:83 miibus0: MII bus on rl0 rlphy0: RealTek internal media interface on miibus0 rlphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto atapci0: HighPoint HPT370 ATA100 controller port 0xa400-0xa4ff,0xa000-0xa003,0x9c00-0x9c07,0x9800-0x9803,0x9400-0x9407 irq 10 at device 6.0 on pci2 ata2: at 0x9400 on atapci0 ata3: at 0x9c00 on atapci0 isab0: Intel 82801BA/BAM (ICH2) PCI to LPC bridge at device 31.0 on pci0 isa0: ISA bus on isab0 atapci1: Intel ICH2 ATA100 controller port 0xf000-0xf00f at device 31.1 on pci0 ata0: at 0x1f0 irq 14 on atapci1 ata1: at 0x170 irq 15 on atapci1 pci0: Intel 82801BA/BAM (ICH2) USB controller USB-A at 31.2 irq 11 pci0: unknown card (vendor=0x8086, dev=0x2443) at 31.3 irq 0 pci0: Intel 82801BA/BAM (ICH2) USB controller USB-B at 31.4 irq 11 pci0: unknown card (vendor=0x8086, dev=0x2445) at 31.5 irq 9 orm0: Option ROM at iomem 0xcc000-0xcf7ff on isa0 fdc0: NEC 72065B or clone at port 0x3f0-0x3f5,0x3f7 irq 6 drq 2 on isa0 fdc0: FIFO enabled, 8 bytes threshold fd0: 1440-KB 3.5 drive on fdc0 drive 0 atkbdc0: Keyboard controller (i8042) at port 0x60,0x64 on isa0 atkbd0: AT Keyboard
Re: [SAGE] FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
Not sure if this is the problem but my copy of MAKEDEV only makes ad0 thru ad3 when you type sh MAKEDEV all You may still need to run sh MAKEDEV ad4 later MEM On Wed, 2003-07-09 at 21:53, Dustin Puryear wrote: Before reading below: I am considering a new install rather than an upgrade of our servers. However, now I just want to beat this problem. :) At 02:19 PM 7/8/2003 +, Phil Pennock wrote: On 2003-07-08 at 14:09 +, Phil Pennock wrote: There was a fairly major update to the IDE disk device handlers which required new device nodes. Bringing in the new MAKEDEV script and running that to create the disk devices will probably let you boot on the new kernel and try to repair things from there. Okay, I'll go get more coffee after sending this. You noticed this one. Sorry. There are two mergemaster steps, one which does the minimum needed to let the rest of the build continue. Yes, I ran 'mergemaster -p' but found that it really only merged my passwd and group files with the new ones. Unless it did some other things behind the scenes then there is no fix here. The issue is, as you note, the device stuff. As I said though, just pull in MAKEDEV manually and run it. That will fix this. Either sh MAKEDEV all or sh MAKEDEV ad0 ... I keep thinking it's the device stuff myself, but even MAKEDEV didn't do the trick (interesting notes below this stuff): # cd /usr/src # make update # /usr/src/usr.sbin/mergemaster/mergemaster.sh -p # make buildworld # make buildkernel # make installkernel # cd /etc # mv MAKEDEV MAKEDEV.old # cp /usr/src/etc/MAKEDEV . # sh MAKEDEV all # reboot Upon reboot I get my favorite lines: blah, blah blah, blah ar0: ... ATA RAID1 array [...] status: READY subdisks: 0 READY ad4: ... at ata2-master UDMA100 1 READY ad6: ... at ata3-master UDMA100 acd0: CDROM ... at ata0-master PIO4 Mounting root from ufs:/dev/ad4s1a Root mount failed: 16 Manual root filesystem specification: blah, blah Okay, so I reboot, but this time I load kernel.good and the system comes up fine with the old 4.4-REL kernel: ar0: ... ATA RAID1 array [...] status: READY subdisks: ad4: ... at ata2-master UDMA100 ad6: ... at ata3-master UDMA100 acd0: CDROM ... at ata0-master using PIO4 Oh, and I found that this server has a HighPoint RAID controller which can do RAID 0, 1, 0+1, and JBOD. Looking in my array configuration I see that we have: ChannelMode Status Primary Master: Mirror (RAID 1) for array #0 UDMA 5HDD0 Secondary Master: Mirror ... UDMA 5Hidden I initially said that this was a Promise controller. Here is my /etc/fstab for /: /dev/ad4s1a / ufs rw 1 1 Following is /var/run/dmesg.log after booting from kernel.good: Copyright (c) 1992-2001 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD 4.4-RELEASE #1: Sun Jun 30 21:23:39 CDT 2002 [EMAIL PROTECTED]:/usr/src/sys/compile/WWW Timecounter i8254 frequency 1193182 Hz Timecounter TSC frequency 1494468860 Hz CPU: Pentium 4 (1494.47-MHz 686-class CPU) Origin = GenuineIntel Id = 0xf07 Stepping = 7 Features=0x3febf9ffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,b28,ACC real memory = 268369920 (262080K bytes) avail memory = 257404928 (251372K bytes) Preloaded elf kernel kernel.good at 0xc0375000. md0: Malloc disk Using $PIR table, 12 entries at 0xc00fdea0 npx0: math processor on motherboard npx0: INT 16 interface pcib0: Host to PCI bridge on motherboard pci0: PCI bus on pcib0 pcib1: PCI to PCI bridge (vendor=8086 device=2532) at device 1.0 on pci0 pci1: PCI bus on pcib1 pci1: NVidia Riva Vanta TNT2 graphics accelerator at 0.0 irq 12 pcib2: Intel 82801BA/BAM (ICH2) Hub to PCI bridge at device 30.0 on pci0 pci2: PCI bus on pcib2 rl0: D-Link DFE-530TX+ 10/100BaseTX port 0x9000-0x90ff mem 0xd900-0xd9ff irq 11 at device 3.0 on pci2 rl0: Ethernet address: 00:50:ba:60:7b:83 miibus0: MII bus on rl0 rlphy0: RealTek internal media interface on miibus0 rlphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto atapci0: HighPoint HPT370 ATA100 controller port 0xa400-0xa4ff,0xa000-0xa003,0x9c00-0x9c07,0x9800-0x9803,0x9400-0x9407 irq 10 at device 6.0 on pci2 ata2: at 0x9400 on atapci0 ata3: at 0x9c00 on atapci0 isab0: Intel 82801BA/BAM (ICH2) PCI to LPC bridge at device 31.0 on pci0 isa0: ISA bus on isab0 atapci1: Intel ICH2 ATA100 controller port 0xf000-0xf00f at device 31.1 on pci0 ata0: at 0x1f0 irq 14 on atapci1 ata1: at 0x170 irq 15 on atapci1 pci0: Intel 82801BA/BAM (ICH2) USB controller USB-A at 31.2 irq 11 pci0: unknown card (vendor=0x8086, dev=0x2443) at 31.3 irq 0 pci0: Intel 82801BA/BAM (ICH2) USB controller USB-B at
Re: FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
Date: Mon, 07 Jul 2003 17:07:39 -0500 From: Dustin Puryear [EMAIL PROTECTED] Sender: [EMAIL PROTECTED] At 02:58 PM 7/7/2003 -0700, Freddie Cash wrote: On July 7, 2003 02:39 pm, you wrote: At 02:27 PM 7/7/2003 -0700, Freddie Cash wrote: On July 7, 2003 02:19 pm, you wrote: There is a blurb near the top that mentions running pwd_mkdb with the -p option to rebuild the master password file. I think this is what is confusing you. My manpage must be outdated or just wrong then: # man mergemaster | grep -- -p the -p option to rebuild your password databases and recreate Read the begining of that sentence. I'll bet it is talking about pwd_mkdb(8) and not mergemaster. Well crap. Still, my version doesn't mention anything about Pre-buildworld mode so the pie on my face is at least tasty. Since you have already updated your source tree, read /usr/src/usr.sbin/mergemaster/mergemaster.8 That is the man page for /usr/src/usr.sbin/mergemaster/mergemaster.sh That will explain everything, including when to run it using the -p parameter. That would be pre-buildworld. Will do. Notice that there is no mention of mergemaster until step 6. The Makefile reads the same on my 4.8 system. However, that should probably be updated, as every time I've seen a mention of mergemaster -p on any of the mailing lists, it has always occured before the buildworld. (And that's the advice I've been following since the -p parameter was introduced.) Yes, updating the docs would be very helpful in situations.. oh, I don't know.. like this one. If you look at the source code (or were here when the -p option was added to mergemaster, you would see that 'mergemaster -p' really just runs mergemaster on a couple of files which could block installworld from working if not already updated. At this time (in either STABLE or CURRENT) it makes absolutely no difference whether it is run before or after build of the world and kernel. It must be run before installworld, although it only makes a difference on limited occasions. (4.4 to 4.8 would be such a case.) FWIW, the only files currently merged when the -p option is present are the group and password files, but this list is subject to change if needed. It is intended to be run before buildworld so that other files can be included down the road if they are needed. But, at this time, running it before installworld is all that is required. -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: [EMAIL PROTECTED] Phone: +1 510 486-8634 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
On July 7, 2003 01:57 pm, Dustin Puryear wrote: I have a test server that is being used to test a 4.4-REL to 4.8-STABLE upgrade, and I've found the system will no longer boot unless I load a saved 4.4-REL kernel. The error I get is along the lines of: Mounting root from ufs:/dev/ad4s1a Root mount failed: 16 Per /usr/src/UPDATING there was a change in the ATA drive at 4.5-REL, but it mentioned running mergemaster to create the new devices, and according to /usr/src/Makefile I shouldn't do that until I get past the current stage (which is where I am stuck): (/usr/src/Makefile comments) 1. cd /usr/src 2. make buildworld 3. make buildkernel 4. make installkernel 5. reboot 6. mergemaster -p There's your problem. You should run mergemaster -p as step 2. The -p means Pre-buildworld. -- Freddie CashDistrict HelpDesk / Network Admin [EMAIL PROTECTED] [EMAIL PROTECTED] (250) 374-0679 ext. 219 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
At 01:59 PM 7/7/2003 -0700, Freddie Cash wrote: (/usr/src/Makefile comments) 1. cd /usr/src 2. make buildworld 3. make buildkernel 4. make installkernel 5. reboot 6. mergemaster -p There's your problem. You should run mergemaster -p as step 2. The -p means Pre-buildworld. This was suggested on another list actually. Unfortunately, either you are wrong or the FreeBSD docs are wrong. In /usr/src/Makefile the steps are listed in the order I show above. Also, per the mergemaster manpage the -p option is used to update the password database and recreate /etc/passwd. Ultimately, I think that this is an issue with the kernel recognizing the storage system, but I could be wrong! Please do correct me if I am. Quick reminder for others: The 4.4-REL kernel is able to mount / fine, but 4.8-STABLE is not. This may either be a symptom of another problem, or the immediate problem. Is this off-topic for freebsd-stable? --- Dustin Puryear [EMAIL PROTECTED] Puryear Information Technology, LLC http://www.puryear-it.com Providing expertise in the management, integration, and security of Windows and UNIX systems, networks, and applications. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
At 02:27 PM 7/7/2003 -0700, Freddie Cash wrote: On July 7, 2003 02:19 pm, you wrote: From the man page for mergemaster: -p Pre-buildworld mode. Compares only files known to be essen- tial to the success of {build|install}world, including /etc/make.conf. There is a blurb near the top that mentions running pwd_mkdb with the -p option to rebuild the master password file. I think this is what is confusing you. My manpage must be outdated or just wrong then: # man mergemaster | grep -- -p the -p option to rebuild your password databases and recreate This is from a longer paragraph, and no mention is made of Pre-buildworld mode. This may be indeed what is throwing me. There seems to also be an error in /usr/src/Makefile since it states that: # For individuals wanting to upgrade their sources (even if only a # delta of a few days): # # 1. `cd /usr/src' (or to the directory containing your source tree). # 2. `make buildworld' # 3. `make buildkernel KERNCONF=YOUR_KERNEL_HERE' (default is GENERIC). # 4. `make installkernel KERNCONF=YOUR_KERNEL_HERE' (default is GENERIC). # 5. `reboot'(in single user mode: boot -s from the loader prompt). # 6. `mergemaster -p' # 7. `make installworld' # 8. `mergemaster' # 9. `reboot' Notice that there is no mention of mergemaster until step 6. This is a test system, but it's remote and I don't want to clobber it unnecessarily. --- Dustin Puryear [EMAIL PROTECTED] Puryear Information Technology, LLC http://www.puryear-it.com Providing expertise in the management, integration, and security of Windows and UNIX systems, networks, and applications. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
At 02:58 PM 7/7/2003 -0700, Freddie Cash wrote: On July 7, 2003 02:39 pm, you wrote: At 02:27 PM 7/7/2003 -0700, Freddie Cash wrote: On July 7, 2003 02:19 pm, you wrote: There is a blurb near the top that mentions running pwd_mkdb with the -p option to rebuild the master password file. I think this is what is confusing you. My manpage must be outdated or just wrong then: # man mergemaster | grep -- -p the -p option to rebuild your password databases and recreate Read the begining of that sentence. I'll bet it is talking about pwd_mkdb(8) and not mergemaster. Well crap. Still, my version doesn't mention anything about Pre-buildworld mode so the pie on my face is at least tasty. Since you have already updated your source tree, read /usr/src/usr.sbin/mergemaster/mergemaster.8 That is the man page for /usr/src/usr.sbin/mergemaster/mergemaster.sh That will explain everything, including when to run it using the -p parameter. That would be pre-buildworld. Will do. Notice that there is no mention of mergemaster until step 6. The Makefile reads the same on my 4.8 system. However, that should probably be updated, as every time I've seen a mention of mergemaster -p on any of the mailing lists, it has always occured before the buildworld. (And that's the advice I've been following since the -p parameter was introduced.) Yes, updating the docs would be very helpful in situations.. oh, I don't know.. like this one. --- Dustin Puryear [EMAIL PROTECTED] Puryear Information Technology, LLC http://www.puryear-it.com Providing expertise in the management, integration, and security of Windows and UNIX systems, networks, and applications. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
FreeBSD 4.4-REL to FreeBSD 4.8-STABLE upgrade problem.
I sent this to freebsd-questions earlier via the newsgroup, but I just found out today that the list-to-news gateway has been down for half a month now. Any idea when it's going back up? Who was hosting that service anyway? And now for the question.. I have a test server that is being used to test a 4.4-REL to 4.8-STABLE upgrade, and I've found the system will no longer boot unless I load a saved 4.4-REL kernel. The error I get is along the lines of: Mounting root from ufs:/dev/ad4s1a Root mount failed: 16 At first I got this error using a custom kernel configuration known as WWW, but after getting this error I built a new kernel based on GENERIC: # cd /usr/src # make update # make buildkernel # make installkernel # reboot Per /usr/src/UPDATING there was a change in the ATA drive at 4.5-REL, but it mentioned running mergemaster to create the new devices, and according to /usr/src/Makefile I shouldn't do that until I get past the current stage (which is where I am stuck): (/usr/src/Makefile comments) 1. cd /usr/src 2. make buildworld 3. make buildkernel 4. make installkernel 5. reboot 6. mergemaster -p This machine does have the Promise RAID-1 IDE controller that is built into a lot motherboards these days. Here is my dmesg output from this server on a successful boot: ar0: 39266MB ATA RAID1 array [5005/255/63] subdisks: ad4: 39266MB IC35L040AVER07-0 [79780/16/63] at ata2-master UDMA100 ad6: 39266MB IC35L040AVER07-0 [79780/16/63] at ata3-master UDMA100 However, this test machine mimics how the web servers were setup, and they do not use /dev/ar, but instead use /dev/ad0: # cat /etc/fstab # DeviceMountpoint FStype Options DumpPass# /dev/ad4s1b noneswapsw 0 0 /dev/ad4s1a / ufs rw 1 1 /dev/ad4s1f /usrufs rw 2 2 /dev/ad4s1e /varufs rw 2 2 /dev/acd0c /cdrom cd9660 ro,noauto 0 0 proc/proc procfs rw 0 0 I am thinking that in 4.5-REL or newer (we are trying to upgrade to 4.8) that things might have changed? Again, /usr/src/UPDATING mentions some changes to the ATA code, but nothing specific to our configuration. Any ideas here? Following are the partition and disklabel information for /dev/ad4 and /dev/ar0: # fdisk /dev/ad4 *** Working on device /dev/ad4 *** parameters extracted from in-core disklabel are: cylinders=5005 heads=255 sectors/track=63 (16065 blks/cyl) Figures below won't work with BIOS for partitions not in cyl 1 parameters to be used for BIOS calculations are: cylinders=5005 heads=255 sectors/track=63 (16065 blks/cyl) Media sector size is 512 Warning: BIOS sector numbering starts with sector 1 Information from DOS bootblock is: The data for partition 1 is: sysid 165,(FreeBSD/NetBSD/386BSD) start 63, size 80405262 (39260 Meg), flag 80 (active) beg: cyl 0/ head 1/ sector 1; end: cyl 1023/ head 254/ sector 63 The data for partition 2 is: UNUSED The data for partition 3 is: UNUSED The data for partition 4 is: UNUSED # disklabel -r ad4s1 # /dev/ad4s1c: type: ESDI disk: ad4s1 label: flags: bytes/sector: 512 sectors/track: 63 tracks/cylinder: 255 sectors/cylinder: 16065 cylinders: 5004 sectors/unit: 80405262 rpm: 3600 interleave: 1 trackskew: 0 cylinderskew: 0 headswitch: 0 # milliseconds track-to-track seek: 0 # milliseconds drivedata: 0 8 partitions: #size offsetfstype [fsize bsize bps/cpg] a: 20480004.2BSD 1024 819216 # (Cyl.0 - 12*) b: 1056032 204800 swap# (Cyl. 12*- 78*) c: 804052620unused0 0 # (Cyl.0 - 5004*) e:40960 12608324.2BSD 1024 819216 # (Cyl. 78*- 81*) f: 79103470 13017924.2BSD 1024 819216 # (Cyl. 81*- 5004*) # fdisk /dev/ar0 *** Working on device /dev/ar0 *** parameters extracted from in-core disklabel are: cylinders=5005 heads=255 sectors/track=63 (16065 blks/cyl) Figures below won't work with BIOS for partitions not in cyl 1 parameters to be used for BIOS calculations are: cylinders=5005 heads=255 sectors/track=63 (16065 blks/cyl) Media sector size is 512 Warning: BIOS sector numbering starts with sector 1 Information from DOS bootblock is: The data for partition 1 is: sysid 165,(FreeBSD/NetBSD/386BSD) start 63, size 80405262 (39260 Meg), flag 80 (active) beg: cyl 0/ head 1/ sector 1; end: cyl 1023/ head 254/ sector 63 The data for partition 2 is: UNUSED The data for partition 3 is: UNUSED The data for partition 4 is: UNUSED # disklabel -r ar0 # /dev/ar0c: type: ESDI disk: ad4s1 label: flags: bytes/sector: 512 sectors/track: 63 tracks/cylinder: 255 sectors/cylinder: 16065 cylinders: 5004 sectors/unit: 80405262 rpm: 3600 interleave: 1 trackskew: 0 cylinderskew: 0 headswitch: 0 # milliseconds track-to-track seek: 0 # milliseconds