alad init on bootup causes panic... - SCO

This is a discussion on alad init on bootup causes panic... - SCO ; I am trying to resurrect an OpenServer 5.0.5 server that has the following configuration... It's on an old Dell PoweEdge server....so it has the 2 embedded SCSI controllers (7890 & 7860 I believe). Those use the blad driver as I ...

+ Reply to Thread
Results 1 to 5 of 5

Thread: alad init on bootup causes panic...

  1. alad init on bootup causes panic...

    I am trying to resurrect an OpenServer 5.0.5 server that has the
    following configuration...

    It's on an old Dell PoweEdge server....so it has the 2 embedded SCSI
    controllers (7890 & 7860 I believe). Those use the blad driver as I
    understand.
    There are 2 mirrored drives that have the OS (mirrored with SCO VDM).

    There is also an Adaptec 2940UW that has 3 drives attached to it.
    (This contains /dev/dsk/vdisk4 - a RAID 5 array that is also RAID'ed
    with SCO VDM)
    THIS is where I need some data - on the /u filesystem that is on
    these 3 disks.

    So, I pull it out of the closet....boot it up.....presto...get a boot
    prompt.

    I hit enter and when it gets to aladinit in the boot up sequence, I
    get a double panic.

    So, I reboot.........and at the boot prompt I enter defbootstr
    disable=alad
    and the system boots up just fine.

    To job my memory, I go look at /etc/conf/cf.d/mscsi - it shows my 3
    disks that are supposed to be on the Adaptec 2940 (all are listed as
    host adapter 0, bus 0, lun 0 and ID's of 0, 1 & 2 - perfect)
    It lists the 2 mirrored OS drives as blad (host adapter 0, bus 0, lun
    0 and ID's of 0 & 1 - perfect)
    Now, this server did use to have a tape drive, so I see it in there as
    well, only it shows to be on alad host adapter 1, bus 0, lun 0 and ID
    of 3 --- uh, okay....
    Also, it shows the cd-rom drive on alad host adapter 2, bus 0, lun 0
    and ID of 5.

    So, this tells me this sytem once upon a time had 3 alad-based SCSI
    controllers at one time, right??
    I look in the PCI slots and there is one PCI slot where the rear cover
    is missing, so you can see to the outside of the case....so, I'm
    guessing someone took a PCI device out of this server at some
    point....?

    Sorry for the long post, but I'm trying to be thorough........

    So, I did mkdev .scsi and removed the tape drive and cd-rom from the
    configuation because I don't need them.....relinked the kernely,
    rebooted and still get double panic...

    My question is this.........what would cause a double panic when
    initializing the adaptec controller? I tried removing the drives from
    the controller, but still panics.....
    I changed the cable, same error..........

    This server isn't on a network so I can't post the exact double panic
    message, but is ends with 0x0000000E at the end....

    If it's absolutely necessary, I can post the full panic message if
    that helps anyone with some insight.....

    Is it possible the kernel is looking for multiple alad based
    controllers and panics when it can only find one??

    HELP!! I would appreciate any insight....I really need to get some
    data off the /u filesystem and this is my last resort.....

    Thanks in advance....

    Scott
    TRG Customer Solutions




  2. Re: alad init on bootup causes panic...


    "scooter" wrote in message
    news:200b50c4-52ac-4749-861d-4a54ab19615c@w7g2000hsa.googlegroups.com...
    >I am trying to resurrect an OpenServer 5.0.5 server that has the
    > following configuration...
    >
    > It's on an old Dell PoweEdge server....so it has the 2 embedded SCSI
    > controllers (7890 & 7860 I believe). Those use the blad driver as I
    > understand.
    > There are 2 mirrored drives that have the OS (mirrored with SCO VDM).
    >
    > There is also an Adaptec 2940UW that has 3 drives attached to it.
    > (This contains /dev/dsk/vdisk4 - a RAID 5 array that is also RAID'ed
    > with SCO VDM)
    > THIS is where I need some data - on the /u filesystem that is on
    > these 3 disks.
    >
    > So, I pull it out of the closet....boot it up.....presto...get a boot
    > prompt.
    >
    > I hit enter and when it gets to aladinit in the boot up sequence, I
    > get a double panic.
    >
    > So, I reboot.........and at the boot prompt I enter defbootstr
    > disable=alad
    > and the system boots up just fine.
    >
    > To job my memory, I go look at /etc/conf/cf.d/mscsi - it shows my 3
    > disks that are supposed to be on the Adaptec 2940 (all are listed as
    > host adapter 0, bus 0, lun 0 and ID's of 0, 1 & 2 - perfect)
    > It lists the 2 mirrored OS drives as blad (host adapter 0, bus 0, lun
    > 0 and ID's of 0 & 1 - perfect)
    > Now, this server did use to have a tape drive, so I see it in there as
    > well, only it shows to be on alad host adapter 1, bus 0, lun 0 and ID
    > of 3 --- uh, okay....
    > Also, it shows the cd-rom drive on alad host adapter 2, bus 0, lun 0
    > and ID of 5.
    >
    > So, this tells me this sytem once upon a time had 3 alad-based SCSI
    > controllers at one time, right??
    > I look in the PCI slots and there is one PCI slot where the rear cover
    > is missing, so you can see to the outside of the case....so, I'm
    > guessing someone took a PCI device out of this server at some
    > point....?
    >
    > Sorry for the long post, but I'm trying to be thorough........
    >
    > So, I did mkdev .scsi and removed the tape drive and cd-rom from the
    > configuation because I don't need them.....relinked the kernely,
    > rebooted and still get double panic...
    >
    > My question is this.........what would cause a double panic when
    > initializing the adaptec controller? I tried removing the drives from
    > the controller, but still panics.....
    > I changed the cable, same error..........
    >
    > This server isn't on a network so I can't post the exact double panic
    > message, but is ends with 0x0000000E at the end....
    >
    > If it's absolutely necessary, I can post the full panic message if
    > that helps anyone with some insight.....
    >
    > Is it possible the kernel is looking for multiple alad based
    > controllers and panics when it can only find one??
    >
    > HELP!! I would appreciate any insight....I really need to get some
    > data off the /u filesystem and this is my last resort.....


    The 0x0E fault is a generic hardware problem. It could very well
    be caused by the missing alad device. It's also possible that by
    removing the 2nd alad adapter, the BIOS has rearranged the IRQ
    assignments and the remaining alad adapter is sharing its IRQ with
    another, non-alad device when it wasn't before.

    Assuming it's the 1st case, boot up the system with disable=alad
    and go into single-user mode. Go to /etc/conf/sdevice.d and edit
    the alad file. Change the "Y" to "N" on the second line to remove
    the second alad device from the kernel. Relink the kernel and
    reboot normally without the disable=alad boot string. It should
    work now.

    If it doesn't, check the scsi bus on the other alad adapter and make
    sure there's an active terminator on the end of the bus. If not, it's
    possible that the last drive on the bus is jumpered to provide
    active termination and term power.

    You can also hit ctrl-A when the 2940UW appears during boot
    to go into its setup mode and check that on board termination
    is enabled there, too.

    Bob


  3. Re: alad init on bootup causes panic...

    On Jul 29, 8:54*pm, "Bob Bailin" <72027.3...@compuserve.com> wrote:
    > "scooter" wrote in message
    >
    > news:200b50c4-52ac-4749-861d-4a54ab19615c@w7g2000hsa.googlegroups.com...
    >
    >
    >
    >
    >
    > >I am trying to resurrect an OpenServer 5.0.5 server that has the
    > > following configuration...

    >
    > > It's on an old Dell PoweEdge server....so it has the 2 embedded SCSI
    > > controllers (7890 & 7860 I believe). *Those use the blad driver as I
    > > understand.
    > > There are 2 mirrored drives that have the OS (mirrored with SCO VDM).

    >
    > > There is also an Adaptec 2940UW that has 3 drives attached to it.
    > > (This contains /dev/dsk/vdisk4 - a RAID 5 array that is also RAID'ed
    > > with SCO VDM)
    > > THIS is where I need some data *- *on the /u *filesystem that is on
    > > these 3 disks.

    >
    > > So, I pull it out of the closet....boot it up.....presto...get a boot
    > > prompt.

    >
    > > I hit enter and when it gets to aladinit in the boot up sequence, I
    > > get a double panic.

    >
    > > So, I reboot.........and at the boot prompt I enter * defbootstr
    > > disable=alad
    > > and the system boots up just fine.

    >
    > > To job my memory, I go look at /etc/conf/cf.d/mscsi - it shows my 3
    > > disks that are supposed to be on the Adaptec 2940 (all are listed as
    > > host adapter 0, bus 0, lun 0 and ID's of 0, 1 & 2 - perfect)
    > > It lists the 2 mirrored OS drives as *blad (host adapter 0, bus 0, lun
    > > 0 and ID's of 0 & 1 - perfect)
    > > Now, this server did use to have a tape drive, so I see it in there as
    > > well, only it shows to be on alad host adapter 1, bus 0, lun 0 and ID
    > > of 3 --- uh, okay....
    > > Also, it shows the cd-rom drive on alad host adapter 2, bus 0, lun 0
    > > and ID of 5.

    >
    > > So, this tells me this sytem once upon a time had 3 alad-based SCSI
    > > controllers at one time, right??
    > > I look in the PCI slots and there is one PCI slot where the rear cover
    > > is missing, so you can see to the outside of the case....so, I'm
    > > guessing someone took a PCI device out of this server at some
    > > point....?

    >
    > > Sorry for the long post, but I'm trying to be thorough........

    >
    > > So, I did mkdev .scsi *and removed the tape drive and cd-rom from the
    > > configuation because I don't need them.....relinked the kernely,
    > > rebooted and still get double panic...

    >
    > > My question is this.........what would cause a double panic when
    > > initializing the adaptec controller? *I tried removing the drives from
    > > the controller, but still panics.....
    > > I changed the cable, same error..........

    >
    > > This server isn't on a network so I can't post the exact double panic
    > > message, but is ends with 0x0000000E at the end....

    >
    > > If it's absolutely necessary, I can post the full panic message if
    > > that helps anyone with some insight.....

    >
    > > Is it possible the kernel is looking for multiple alad based
    > > controllers and panics when it can only find one??

    >
    > > HELP!! *I would appreciate any insight....I really need to get some
    > > data off the /u filesystem and this is my last resort.....

    >
    > The 0x0E fault is a generic hardware problem. It could very well
    > be caused by the missing alad device. It's also possible that by
    > removing the 2nd alad adapter, the BIOS has rearranged the IRQ
    > assignments and the remaining alad adapter is sharing its IRQ with
    > another, non-alad device when it wasn't before.
    >
    > Assuming it's the 1st case, boot up the system with disable=alad
    > and go into single-user mode. Go to /etc/conf/sdevice.d and edit
    > the alad file. Change the "Y" to "N" on the second line to remove
    > the second alad device from the kernel. Relink the kernel and
    > reboot normally without the disable=alad boot string. It should
    > work now.
    >
    > If it doesn't, check the scsi bus on the other alad adapter and make
    > sure there's an active terminator on the end of the bus. If not, it's
    > possible that the last drive on the bus is jumpered to provide
    > active termination and term power.
    >
    > You can also hit ctrl-A when the 2940UW appears during boot
    > to go into its setup mode and check that on board termination
    > is enabled there, too.
    >
    > Bob- Hide quoted text -
    >
    > - Show quoted text -


    Thanks Bob -

    That didn't work........but I started looking into possible IRQ
    conflicts, etc.
    Since this server hadn't been booted in a while, I had to go into
    setup to set the time, etc so I guess it restored the defaults in the
    system BIOS....so, I disabled the 2nd SCSI on-board adapter (since I
    have nothing hooked to it) and voila! It boots past the alad init
    message, etc.........then after hitting Ctrl-D and entering thru the
    time, it starts to load /dev/dsk/vdisk4 (RAID 5) and it
    panics........ughhh.......so, I'm guessing perhaps one of the drives
    (or more) went bad? I would figure though if that was the case, it
    would simply say it couldn't load /dev/dsk/vdisk4 because it was out
    of parity or dead drives, etc....I wouldn't figure a failed drive
    would cause a kernel panic......?? Is that a correct assumption on my
    part?

    I will check SCSI termination in the morning, but I have no reason to
    believe that's the problem since there is a terminator on the end of
    the SCSI cable and it's the same cable that was in it before, etc.

    Any other thoughts, ideas, suggestions? for my desperate self??? LOL

    Ughhh....if.....I......could......just........get. ............it................to.................. ........boot............ /
    u ........!!!

    Thanks again for any insight.....

    Scott

  4. Re: alad init on bootup causes panic...

    Sorry, nothing more to add. I've never been a fan of software RAID,
    precisely for the problems you're having now. Perhaps if you post
    the panic you get attempting to mount the RAID volume, someone
    more familiar with SCO VDM might chime in.

    "scooter" wrote in message
    news:f0ae4350-bc1a-402c-898b-8beab7ac35c7@r66g2000hsg.googlegroups.com...
    On Jul 29, 8:54 pm, "Bob Bailin" <72027.3...@compuserve.com> wrote:
    > "scooter" wrote in message
    >
    > news:200b50c4-52ac-4749-861d-4a54ab19615c@w7g2000hsa.googlegroups.com...
    >
    >
    >
    >
    >
    > >I am trying to resurrect an OpenServer 5.0.5 server that has the
    > > following configuration...

    >
    > > It's on an old Dell PoweEdge server....so it has the 2 embedded SCSI
    > > controllers (7890 & 7860 I believe). Those use the blad driver as I
    > > understand.
    > > There are 2 mirrored drives that have the OS (mirrored with SCO VDM).

    >
    > > There is also an Adaptec 2940UW that has 3 drives attached to it.
    > > (This contains /dev/dsk/vdisk4 - a RAID 5 array that is also RAID'ed
    > > with SCO VDM)
    > > THIS is where I need some data - on the /u filesystem that is on
    > > these 3 disks.

    >
    > > So, I pull it out of the closet....boot it up.....presto...get a boot
    > > prompt.

    >
    > > I hit enter and when it gets to aladinit in the boot up sequence, I
    > > get a double panic.

    >
    > > So, I reboot.........and at the boot prompt I enter defbootstr
    > > disable=alad
    > > and the system boots up just fine.

    >
    > > To job my memory, I go look at /etc/conf/cf.d/mscsi - it shows my 3
    > > disks that are supposed to be on the Adaptec 2940 (all are listed as
    > > host adapter 0, bus 0, lun 0 and ID's of 0, 1 & 2 - perfect)
    > > It lists the 2 mirrored OS drives as blad (host adapter 0, bus 0, lun
    > > 0 and ID's of 0 & 1 - perfect)
    > > Now, this server did use to have a tape drive, so I see it in there as
    > > well, only it shows to be on alad host adapter 1, bus 0, lun 0 and ID
    > > of 3 --- uh, okay....
    > > Also, it shows the cd-rom drive on alad host adapter 2, bus 0, lun 0
    > > and ID of 5.

    >
    > > So, this tells me this sytem once upon a time had 3 alad-based SCSI
    > > controllers at one time, right??
    > > I look in the PCI slots and there is one PCI slot where the rear cover
    > > is missing, so you can see to the outside of the case....so, I'm
    > > guessing someone took a PCI device out of this server at some
    > > point....?

    >
    > > Sorry for the long post, but I'm trying to be thorough........

    >
    > > So, I did mkdev .scsi and removed the tape drive and cd-rom from the
    > > configuation because I don't need them.....relinked the kernely,
    > > rebooted and still get double panic...

    >
    > > My question is this.........what would cause a double panic when
    > > initializing the adaptec controller? I tried removing the drives from
    > > the controller, but still panics.....
    > > I changed the cable, same error..........

    >
    > > This server isn't on a network so I can't post the exact double panic
    > > message, but is ends with 0x0000000E at the end....

    >
    > > If it's absolutely necessary, I can post the full panic message if
    > > that helps anyone with some insight.....

    >
    > > Is it possible the kernel is looking for multiple alad based
    > > controllers and panics when it can only find one??

    >
    > > HELP!! I would appreciate any insight....I really need to get some
    > > data off the /u filesystem and this is my last resort.....

    >
    > The 0x0E fault is a generic hardware problem. It could very well
    > be caused by the missing alad device. It's also possible that by
    > removing the 2nd alad adapter, the BIOS has rearranged the IRQ
    > assignments and the remaining alad adapter is sharing its IRQ with
    > another, non-alad device when it wasn't before.
    >
    > Assuming it's the 1st case, boot up the system with disable=alad
    > and go into single-user mode. Go to /etc/conf/sdevice.d and edit
    > the alad file. Change the "Y" to "N" on the second line to remove
    > the second alad device from the kernel. Relink the kernel and
    > reboot normally without the disable=alad boot string. It should
    > work now.
    >
    > If it doesn't, check the scsi bus on the other alad adapter and make
    > sure there's an active terminator on the end of the bus. If not, it's
    > possible that the last drive on the bus is jumpered to provide
    > active termination and term power.
    >
    > You can also hit ctrl-A when the 2940UW appears during boot
    > to go into its setup mode and check that on board termination
    > is enabled there, too.
    >
    > Bob- Hide quoted text -
    >
    > - Show quoted text -


    Thanks Bob -

    That didn't work........but I started looking into possible IRQ
    conflicts, etc.
    Since this server hadn't been booted in a while, I had to go into
    setup to set the time, etc so I guess it restored the defaults in the
    system BIOS....so, I disabled the 2nd SCSI on-board adapter (since I
    have nothing hooked to it) and voila! It boots past the alad init
    message, etc.........then after hitting Ctrl-D and entering thru the
    time, it starts to load /dev/dsk/vdisk4 (RAID 5) and it
    panics........ughhh.......so, I'm guessing perhaps one of the drives
    (or more) went bad? I would figure though if that was the case, it
    would simply say it couldn't load /dev/dsk/vdisk4 because it was out
    of parity or dead drives, etc....I wouldn't figure a failed drive
    would cause a kernel panic......?? Is that a correct assumption on my
    part?

    I will check SCSI termination in the morning, but I have no reason to
    believe that's the problem since there is a terminator on the end of
    the SCSI cable and it's the same cable that was in it before, etc.

    Any other thoughts, ideas, suggestions? for my desperate self??? LOL

    Ughhh....if.....I......could......just........get. ............it...............to................... .......boot............
    /
    u ........!!!

    Thanks again for any insight.....

    Scott


  5. Re: alad init on bootup causes panic...

    On Jul 30, 10:14*pm, "Bob Bailin" <72027.3...@compuserve.com> wrote:
    > Sorry, nothing more to add. I've never been a fan of software RAID,
    > precisely for the problems you're having now. Perhaps if you post
    > the panic you get attempting to mount the RAID volume, someone
    > more familiar with SCO VDM might chime in.
    >
    > "scooter" wrote in message
    >
    > news:f0ae4350-bc1a-402c-898b-8beab7ac35c7@r66g2000hsg.googlegroups.com...
    > On Jul 29, 8:54 pm, "Bob Bailin" <72027.3...@compuserve.com> wrote:
    >
    >
    >
    >
    >
    > > "scooter" wrote in message

    >
    > >news:200b50c4-52ac-4749-861d-4a54ab19615c@w7g2000hsa.googlegroups.com...

    >
    > > >I am trying to resurrect an OpenServer 5.0.5 server that has the
    > > > following configuration...

    >
    > > > It's on an old Dell PoweEdge server....so it has the 2 embedded SCSI
    > > > controllers (7890 & 7860 I believe). Those use the blad driver as I
    > > > understand.
    > > > There are 2 mirrored drives that have the OS (mirrored with SCO VDM).

    >
    > > > There is also an Adaptec 2940UW that has 3 drives attached to it.
    > > > (This contains /dev/dsk/vdisk4 - a RAID 5 array that is also RAID'ed
    > > > with SCO VDM)
    > > > THIS is where I need some data - on the /u filesystem that is on
    > > > these 3 disks.

    >
    > > > So, I pull it out of the closet....boot it up.....presto...get a boot
    > > > prompt.

    >
    > > > I hit enter and when it gets to aladinit in the boot up sequence, I
    > > > get a double panic.

    >
    > > > So, I reboot.........and at the boot prompt I enter defbootstr
    > > > disable=alad
    > > > and the system boots up just fine.

    >
    > > > To job my memory, I go look at /etc/conf/cf.d/mscsi - it shows my 3
    > > > disks that are supposed to be on the Adaptec 2940 (all are listed as
    > > > host adapter 0, bus 0, lun 0 and ID's of 0, 1 & 2 - perfect)
    > > > It lists the 2 mirrored OS drives as blad (host adapter 0, bus 0, lun
    > > > 0 and ID's of 0 & 1 - perfect)
    > > > Now, this server did use to have a tape drive, so I see it in there as
    > > > well, only it shows to be on alad host adapter 1, bus 0, lun 0 and ID
    > > > of 3 --- uh, okay....
    > > > Also, it shows the cd-rom drive on alad host adapter 2, bus 0, lun 0
    > > > and ID of 5.

    >
    > > > So, this tells me this sytem once upon a time had 3 alad-based SCSI
    > > > controllers at one time, right??
    > > > I look in the PCI slots and there is one PCI slot where the rear cover
    > > > is missing, so you can see to the outside of the case....so, I'm
    > > > guessing someone took a PCI device out of this server at some
    > > > point....?

    >
    > > > Sorry for the long post, but I'm trying to be thorough........

    >
    > > > So, I did mkdev .scsi and removed the tape drive and cd-rom from the
    > > > configuation because I don't need them.....relinked the kernely,
    > > > rebooted and still get double panic...

    >
    > > > My question is this.........what would cause a double panic when
    > > > initializing the adaptec controller? I tried removing the drives from
    > > > the controller, but still panics.....
    > > > I changed the cable, same error..........

    >
    > > > This server isn't on a network so I can't post the exact double panic
    > > > message, but is ends with 0x0000000E at the end....

    >
    > > > If it's absolutely necessary, I can post the full panic message if
    > > > that helps anyone with some insight.....

    >
    > > > Is it possible the kernel is looking for multiple alad based
    > > > controllers and panics when it can only find one??

    >
    > > > HELP!! I would appreciate any insight....I really need to get some
    > > > data off the /u filesystem and this is my last resort.....

    >
    > > The 0x0E fault is a generic hardware problem. It could very well
    > > be caused by the missing alad device. It's also possible that by
    > > removing the 2nd alad adapter, the BIOS has rearranged the IRQ
    > > assignments and the remaining alad adapter is sharing its IRQ with
    > > another, non-alad device when it wasn't before.

    >
    > > Assuming it's the 1st case, boot up the system with disable=alad
    > > and go into single-user mode. Go to /etc/conf/sdevice.d and edit
    > > the alad file. Change the "Y" to "N" on the second line to remove
    > > the second alad device from the kernel. Relink the kernel and
    > > reboot normally without the disable=alad boot string. It should
    > > work now.

    >
    > > If it doesn't, check the scsi bus on the other alad adapter and make
    > > sure there's an active terminator on the end of the bus. If not, it's
    > > possible that the last drive on the bus is jumpered to provide
    > > active termination and term power.

    >
    > > You can also hit ctrl-A when the 2940UW appears during boot
    > > to go into its setup mode and check that on board termination
    > > is enabled there, too.

    >
    > > Bob- Hide quoted text -

    >
    > > - Show quoted text -

    >
    > Thanks Bob -
    >
    > That didn't work........but I started looking into possible IRQ
    > conflicts, etc.
    > Since this server hadn't been booted in a while, I had to go into
    > setup to set the time, etc so I guess it restored the defaults in the
    > system BIOS....so, I disabled the 2nd SCSI on-board adapter (since I
    > have nothing hooked to it) and voila! It boots past the alad init
    > message, etc.........then after hitting Ctrl-D and entering thru the
    > time, it starts to load /dev/dsk/vdisk4 (RAID 5) and it
    > panics........ughhh.......so, I'm guessing perhaps one of the drives
    > (or more) went bad? *I would figure though if that was the case, it
    > would simply say it couldn't load /dev/dsk/vdisk4 because it was out
    > of parity or dead drives, etc....I wouldn't figure a failed drive
    > would cause a kernel panic......?? *Is that a correct assumption on my
    > part?
    >
    > I will check SCSI termination in the morning, but I have no reason to
    > believe that's the problem since there is a terminator on the end of
    > the SCSI cable and it's the same cable that was in it before, etc.
    >
    > Any other thoughts, ideas, suggestions? for my desperate self??? LOL
    >
    > Ughhh....if.....I......could......just........get. ............it............*....to................. .........boot............
    > /
    > u ........!!!
    >
    > Thanks again for any insight.....
    >
    > Scott- Hide quoted text -
    >
    > - Show quoted text -



    Thanks Bob for your input.........

    I ultimately disconnected the first drive in the RAID...the system
    squawked of course...but it booted and mounted the /u filesystem and I
    got what I needed off of it....

    thanks again

    Scott

+ Reply to Thread