CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk - SCO

This is a discussion on CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk - SCO ; Can anyone help me with this? Dual-Processor Dell Poweredge running SCO 5.0.6 Patches - RS506A Up until recently this server has been running fine until the air con unit leaked onto it. This caused a problem with one of the ...

+ Reply to Thread
Results 1 to 12 of 12

Thread: CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

  1. CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    Can anyone help me with this?

    Dual-Processor Dell Poweredge running SCO 5.0.6
    Patches - RS506A

    Up until recently this server has been running fine until the air con
    unit leaked onto it. This caused a problem with one of the PSU's.
    After switching the server off, and leaving the cover off it overnight
    we were then able to power up the server.
    It lasted about 2hrs before the users were getting errors in the
    software and then 3 disks started to beep. When rebooting the server
    it came up with 1 logical drive failed.

    There are 6 disks in this server
    1 x mirrored pair
    1 x Raid 5

    3 of the disks in the raid 5 were flashing. Hence it would boot to the
    SCO login prompt but wouldn't mount the data array.
    We were able to go into the SCSI utilities (CTRL M) and select these 3
    disks and force them on-line. The beeping & flashing went away and
    were were now able to mount the data array. Users were able to log back
    on - all seemed fine - no further errors reported.

    The above happened on a Friday and then on the Monday morning the
    server panicked with the following:

    PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd (1/104)
    cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages .
    Dump not complete .

    When server was rebooted it said that there were no logical drives on
    the host adapter and 0 logical drives handled by the bios. Hardware
    engineer went to site to replace the PERC card - tried to replace this
    but it then wouldn't see the O/S disks. He then put the old card back
    in place and managed to get it working.

    The server has been up and running for a couple of weeks with no
    problems apart from every now and again I get the following error
    message written to syslog and messages:

    CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    (ha=0 bus=0 id=0 lun=1) block=11066764

    It is not always the same block number listed.
    Why does this error only occur for CPU2?
    How do I rectify this problem?

    Any help would be greatly appreciated

    Sarah


  2. Re: CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    In article <1154702127.932930.100950@h48g2000cwc.googlegroups. com>,
    wrote:
    >Can anyone help me with this?
    >
    >Dual-Processor Dell Poweredge running SCO 5.0.6
    >Patches - RS506A
    >
    >Up until recently this server has been running fine until the air con
    >unit leaked onto it. This caused a problem with one of the PSU's.
    >After switching the server off, and leaving the cover off it overnight
    >we were then able to power up the server.
    >It lasted about 2hrs before the users were getting errors in the
    >software and then 3 disks started to beep. When rebooting the server
    >it came up with 1 logical drive failed.
    >
    >There are 6 disks in this server
    >1 x mirrored pair
    >1 x Raid 5
    >
    >3 of the disks in the raid 5 were flashing. Hence it would boot to the
    >SCO login prompt but wouldn't mount the data array.
    >We were able to go into the SCSI utilities (CTRL M) and select these 3
    >disks and force them on-line. The beeping & flashing went away and
    >were were now able to mount the data array. Users were able to log back
    >on - all seemed fine - no further errors reported.
    >
    >The above happened on a Friday and then on the Monday morning the
    >server panicked with the following:
    >
    >PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd (1/104)
    >cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages .
    >Dump not complete .
    >
    >When server was rebooted it said that there were no logical drives on
    >the host adapter and 0 logical drives handled by the bios. Hardware
    >engineer went to site to replace the PERC card - tried to replace this
    >but it then wouldn't see the O/S disks. He then put the old card back
    >in place and managed to get it working.
    >
    >The server has been up and running for a couple of weeks with no
    >problems apart from every now and again I get the following error
    >message written to syslog and messages:
    >
    >CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    >(ha=0 bus=0 id=0 lun=1) block=11066764
    >
    >It is not always the same block number listed.
    >Why does this error only occur for CPU2?
    >How do I rectify this problem?
    >
    >Any help would be greatly appreciated
    >
    >Sarah


    Having seen moisture get into computers before sometimes you just
    need to start over with a fresh motherboard, and perhaps
    controllers.

    Since it ran find before and now just runs a day or so before you
    have the problem I suspect there is some hardware failure.

    Moisture can cause shorts - and perhaps something was stressed
    enought to make it unreliable, and will run a few hours/days before
    something makes it fail and give errors.

    Water will NOT hurt computers/electonics IF it is turned off and
    then let dry thoroughly. Of course that pre-supposes 'clean'
    water.

    A very long time ago when I worked in broadcast the station I was
    at bought an FM transmitter that had been put on line in the late
    1940s when FM was 'the coming thing'. It turned out FM
    practically died except in major cities.

    The transmitter sat in a chicken coop for over 10 years.

    It was covered with a lot of chicken excrement.

    The solution was to take a hose and thoroughly wash everything, and
    then let it set and complete dry out for about a week.

    But moisture on a running piece of equipment usually means you can
    not trust that equipment any more - at least in critical
    applications.

    Bill


    --
    Bill Vermillion - bv @ wjv . com

  3. Re: CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    sarahbeattie101@yahoo.co.uk wrote:
    > Can anyone help me with this?
    >
    > Dual-Processor Dell Poweredge running SCO 5.0.6
    > Patches - RS506A
    >
    > Up until recently this server has been running fine until the air con
    > unit leaked onto it. This caused a problem with one of the PSU's.
    > After switching the server off, and leaving the cover off it overnight
    > we were then able to power up the server.
    > It lasted about 2hrs before the users were getting errors in the
    > software and then 3 disks started to beep. When rebooting the server
    > it came up with 1 logical drive failed.
    >
    > There are 6 disks in this server
    > 1 x mirrored pair
    > 1 x Raid 5
    >
    > 3 of the disks in the raid 5 were flashing. Hence it would boot to the
    > SCO login prompt but wouldn't mount the data array.
    > We were able to go into the SCSI utilities (CTRL M) and select these 3
    > disks and force them on-line. The beeping & flashing went away and
    > were were now able to mount the data array. Users were able to log back
    > on - all seemed fine - no further errors reported.
    >
    > The above happened on a Friday and then on the Monday morning the
    > server panicked with the following:
    >
    > PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd (1/104)
    > cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages .
    > Dump not complete .
    >
    > When server was rebooted it said that there were no logical drives on
    > the host adapter and 0 logical drives handled by the bios. Hardware
    > engineer went to site to replace the PERC card - tried to replace this
    > but it then wouldn't see the O/S disks. He then put the old card back
    > in place and managed to get it working.
    >
    > The server has been up and running for a couple of weeks with no
    > problems apart from every now and again I get the following error
    > message written to syslog and messages:
    >
    > CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    > (ha=0 bus=0 id=0 lun=1) block=11066764
    >
    > It is not always the same block number listed.
    > Why does this error only occur for CPU2?
    > How do I rectify this problem?
    >
    > Any help would be greatly appreciated
    >
    > Sarah
    >


    How old is the server? How important is your data to you?

    That machine is obviously unreliable at this point. Is it still under
    warranty? Does Dell cover things like moisture incursion?

    If so I would be putting pressure on the to replace EVERYTHING. If it's
    no longer under warranty or covered - think real hard about getting a
    new one.

    That message indicates unrecoverable hard disk errors when data is read
    or written - you are losing data.

    I hope you are getting verified backups via Backup Edge or Lone-tar. I
    would not trust backups via tar or cpio.

    --
    ----------------------------------------------------
    Pat Welch, UBB Computer Services, a WCS Affiliate
    SCO Authorized Partner
    Unix/Linux/Windows/Hardware Sales/Support
    (209) 745-1401 Cell: (209) 251-9120
    E-mail: patubb@inreach.com
    ----------------------------------------------------

  4. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk


    wrote in message
    news:1154702127.932930.100950@h48g2000cwc.googlegr oups.com...
    > Can anyone help me with this?
    >
    > Dual-Processor Dell Poweredge running SCO 5.0.6
    > Patches - RS506A
    >
    > Up until recently this server has been running fine until the air con
    > unit leaked onto it. This caused a problem with one of the PSU's.
    > After switching the server off, and leaving the cover off it overnight
    > we were then able to power up the server.
    > It lasted about 2hrs before the users were getting errors in the
    > software and then 3 disks started to beep. When rebooting the server
    > it came up with 1 logical drive failed.
    >
    > There are 6 disks in this server
    > 1 x mirrored pair
    > 1 x Raid 5
    >
    > 3 of the disks in the raid 5 were flashing. Hence it would boot to the
    > SCO login prompt but wouldn't mount the data array.
    > We were able to go into the SCSI utilities (CTRL M) and select these 3
    > disks and force them on-line. The beeping & flashing went away and
    > were were now able to mount the data array. Users were able to log back
    > on - all seemed fine - no further errors reported.
    >
    > The above happened on a Friday and then on the Monday morning the
    > server panicked with the following:
    >
    > PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd (1/104)
    > cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages .
    > Dump not complete .
    >
    > When server was rebooted it said that there were no logical drives on
    > the host adapter and 0 logical drives handled by the bios. Hardware
    > engineer went to site to replace the PERC card - tried to replace this
    > but it then wouldn't see the O/S disks. He then put the old card back
    > in place and managed to get it working.
    >
    > The server has been up and running for a couple of weeks with no
    > problems apart from every now and again I get the following error
    > message written to syslog and messages:
    >
    > CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    > (ha=0 bus=0 id=0 lun=1) block=11066764
    >
    > It is not always the same block number listed.
    > Why does this error only occur for CPU2?
    > How do I rectify this problem?
    >
    > Any help would be greatly appreciated
    >
    > Sarah
    >


    Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    issues, just shut it down, pull every stinking board that is plugged in, dry
    them by wiping, blowing, and in some instances using alcohol to make sure
    that any and all visible water has been dried. Then use the old hair drier
    to blow the port slots dry. Once this is done, let it sit for at least 24
    hours with a fan blowing on it. Now you can re-insert the boards, boot it
    up, make a good backup, shut it down, and replace the whole thing with a new
    machine.

    The warning is this. WATER conducts electriciy and infiltrates in ways we
    can not predict. Therefore, even by drying, we can not be sure that the
    water has not caused unseen damage. When dealing with data integrity,
    customer files, security related information, we can not and should not ever
    assume things will be OK. We must KNOW they will. Thus, replacement of
    equipment is the only reasonable answer.

    I could tell numerous horror stories of water damaged machines (I've been
    involved with many), and none of them ends in story-book fashion.

    I hope you have good backups. Good luck on the recovery.

    JP - piperent



  5. Re: CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    Hi Pat

    The server is probably just over a year old - not sure what the
    warranty is on it - will check that out.
    Why would only CPU2 report the errors?
    Do you know of any diagnostics that can be run to determine whether
    CPU2 is having problems?
    I believe that this customer does have insurance for this type of thing
    and so will discuss this issue with him

    Thanks for your reply
    Sarah


    Pat Welch wrote:
    > sarahbeattie101@yahoo.co.uk wrote:
    > > Can anyone help me with this?
    > >
    > > Dual-Processor Dell Poweredge running SCO 5.0.6
    > > Patches - RS506A
    > >
    > > Up until recently this server has been running fine until the air con
    > > unit leaked onto it. This caused a problem with one of the PSU's.
    > > After switching the server off, and leaving the cover off it overnight
    > > we were then able to power up the server.
    > > It lasted about 2hrs before the users were getting errors in the
    > > software and then 3 disks started to beep. When rebooting the server
    > > it came up with 1 logical drive failed.
    > >
    > > There are 6 disks in this server
    > > 1 x mirrored pair
    > > 1 x Raid 5
    > >
    > > 3 of the disks in the raid 5 were flashing. Hence it would boot to the
    > > SCO login prompt but wouldn't mount the data array.
    > > We were able to go into the SCSI utilities (CTRL M) and select these 3
    > > disks and force them on-line. The beeping & flashing went away and
    > > were were now able to mount the data array. Users were able to log back
    > > on - all seemed fine - no further errors reported.
    > >
    > > The above happened on a Friday and then on the Monday morning the
    > > server panicked with the following:
    > >
    > > PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd (1/104)
    > > cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages .
    > > Dump not complete .
    > >
    > > When server was rebooted it said that there were no logical drives on
    > > the host adapter and 0 logical drives handled by the bios. Hardware
    > > engineer went to site to replace the PERC card - tried to replace this
    > > but it then wouldn't see the O/S disks. He then put the old card back
    > > in place and managed to get it working.
    > >
    > > The server has been up and running for a couple of weeks with no
    > > problems apart from every now and again I get the following error
    > > message written to syslog and messages:
    > >
    > > CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    > > (ha=0 bus=0 id=0 lun=1) block=11066764
    > >
    > > It is not always the same block number listed.
    > > Why does this error only occur for CPU2?
    > > How do I rectify this problem?
    > >
    > > Any help would be greatly appreciated
    > >
    > > Sarah
    > >

    >
    > How old is the server? How important is your data to you?
    >
    > That machine is obviously unreliable at this point. Is it still under
    > warranty? Does Dell cover things like moisture incursion?
    >
    > If so I would be putting pressure on the to replace EVERYTHING. If it's
    > no longer under warranty or covered - think real hard about getting a
    > new one.
    >
    > That message indicates unrecoverable hard disk errors when data is read
    > or written - you are losing data.
    >
    > I hope you are getting verified backups via Backup Edge or Lone-tar. I
    > would not trust backups via tar or cpio.
    >
    > --
    > ----------------------------------------------------
    > Pat Welch, UBB Computer Services, a WCS Affiliate
    > SCO Authorized Partner
    > Unix/Linux/Windows/Hardware Sales/Support
    > (209) 745-1401 Cell: (209) 251-9120
    > E-mail: patubb@inreach.com
    > ----------------------------------------------------



  6. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    Thanks JP

    Will see whether it is still under warranty with Dell on Monday.
    If not customer said that he was insured.
    Will get this sorted sooner rather than later as I don't want to go
    through the pain of this again.

    Sarah

    JP wrote:
    > wrote in message
    > news:1154702127.932930.100950@h48g2000cwc.googlegr oups.com...
    > > Can anyone help me with this?
    > >
    > > Dual-Processor Dell Poweredge running SCO 5.0.6
    > > Patches - RS506A
    > >
    > > Up until recently this server has been running fine until the air con
    > > unit leaked onto it. This caused a problem with one of the PSU's.
    > > After switching the server off, and leaving the cover off it overnight
    > > we were then able to power up the server.
    > > It lasted about 2hrs before the users were getting errors in the
    > > software and then 3 disks started to beep. When rebooting the server
    > > it came up with 1 logical drive failed.
    > >
    > > There are 6 disks in this server
    > > 1 x mirrored pair
    > > 1 x Raid 5
    > >
    > > 3 of the disks in the raid 5 were flashing. Hence it would boot to the
    > > SCO login prompt but wouldn't mount the data array.
    > > We were able to go into the SCSI utilities (CTRL M) and select these 3
    > > disks and force them on-line. The beeping & flashing went away and
    > > were were now able to mount the data array. Users were able to log back
    > > on - all seemed fine - no further errors reported.
    > >
    > > The above happened on a Friday and then on the Monday morning the
    > > server panicked with the following:
    > >
    > > PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd (1/104)
    > > cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages .
    > > Dump not complete .
    > >
    > > When server was rebooted it said that there were no logical drives on
    > > the host adapter and 0 logical drives handled by the bios. Hardware
    > > engineer went to site to replace the PERC card - tried to replace this
    > > but it then wouldn't see the O/S disks. He then put the old card back
    > > in place and managed to get it working.
    > >
    > > The server has been up and running for a couple of weeks with no
    > > problems apart from every now and again I get the following error
    > > message written to syslog and messages:
    > >
    > > CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    > > (ha=0 bus=0 id=0 lun=1) block=11066764
    > >
    > > It is not always the same block number listed.
    > > Why does this error only occur for CPU2?
    > > How do I rectify this problem?
    > >
    > > Any help would be greatly appreciated
    > >
    > > Sarah
    > >

    >
    > Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    > issues, just shut it down, pull every stinking board that is plugged in, dry
    > them by wiping, blowing, and in some instances using alcohol to make sure
    > that any and all visible water has been dried. Then use the old hair drier
    > to blow the port slots dry. Once this is done, let it sit for at least 24
    > hours with a fan blowing on it. Now you can re-insert the boards, boot it
    > up, make a good backup, shut it down, and replace the whole thing with a new
    > machine.
    >
    > The warning is this. WATER conducts electriciy and infiltrates in ways we
    > can not predict. Therefore, even by drying, we can not be sure that the
    > water has not caused unseen damage. When dealing with data integrity,
    > customer files, security related information, we can not and should not ever
    > assume things will be OK. We must KNOW they will. Thus, replacement of
    > equipment is the only reasonable answer.
    >
    > I could tell numerous horror stories of water damaged machines (I've been
    > involved with many), and none of them ends in story-book fashion.
    >
    > I hope you have good backups. Good luck on the recovery.
    >
    > JP - piperent



  7. Re: CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk


    wrote in message
    news:1154794011.415043.230820@b28g2000cwb.googlegr oups.com...
    > Hi Pat
    >
    > The server is probably just over a year old - not sure what the
    > warranty is on it - will check that out.
    > Why would only CPU2 report the errors?
    > Do you know of any diagnostics that can be run to determine whether
    > CPU2 is having problems?
    > I believe that this customer does have insurance for this type of thing
    > and so will discuss this issue with him
    >
    > Thanks for your reply
    > Sarah


    Use displayintr to show which cpu's are handling which hardware driver
    interrupts. Depending on how the driver is written, the interrupt may
    be fixed to a specific cpu, shareable among all cpu's, or be assignable
    at boot time.

    See man displayintr (ADM) for more details.

    Bob

    >
    >
    > Pat Welch wrote:
    > > sarahbeattie101@yahoo.co.uk wrote:
    > > > Can anyone help me with this?
    > > >
    > > > Dual-Processor Dell Poweredge running SCO 5.0.6
    > > > Patches - RS506A
    > > >
    > > > Up until recently this server has been running fine until the air con
    > > > unit leaked onto it. This caused a problem with one of the PSU's.
    > > > After switching the server off, and leaving the cover off it overnight
    > > > we were then able to power up the server.
    > > > It lasted about 2hrs before the users were getting errors in the
    > > > software and then 3 disks started to beep. When rebooting the server
    > > > it came up with 1 logical drive failed.
    > > >
    > > > There are 6 disks in this server
    > > > 1 x mirrored pair
    > > > 1 x Raid 5
    > > >
    > > > 3 of the disks in the raid 5 were flashing. Hence it would boot to

    the
    > > > SCO login prompt but wouldn't mount the data array.
    > > > We were able to go into the SCSI utilities (CTRL M) and select these 3
    > > > disks and force them on-line. The beeping & flashing went away and
    > > > were were now able to mount the data array. Users were able to log

    back
    > > > on - all seemed fine - no further errors reported.
    > > >
    > > > The above happened on a Friday and then on the Monday morning the
    > > > server panicked with the following:
    > > >
    > > > PANIC:HTFS block 119110532 epi already allocated on HTFS dev hd

    (1/104)
    > > > cannot dump 524127 pages th dump hd(1/41) space for only 131072 pages

    ..
    > > > Dump not complete .
    > > >
    > > > When server was rebooted it said that there were no logical drives on
    > > > the host adapter and 0 logical drives handled by the bios. Hardware
    > > > engineer went to site to replace the PERC card - tried to replace this
    > > > but it then wouldn't see the O/S disks. He then put the old card back
    > > > in place and managed to get it working.
    > > >
    > > > The server has been up and running for a couple of weeks with no
    > > > problems apart from every now and again I get the following error
    > > > message written to syslog and messages:
    > > >
    > > > CPU2: NOTICE: Sdsk: Unrecoverable error writing SCSI disk 1 dev 1/104
    > > > (ha=0 bus=0 id=0 lun=1) block=11066764
    > > >
    > > > It is not always the same block number listed.
    > > > Why does this error only occur for CPU2?
    > > > How do I rectify this problem?
    > > >
    > > > Any help would be greatly appreciated
    > > >
    > > > Sarah
    > > >

    > >
    > > How old is the server? How important is your data to you?
    > >
    > > That machine is obviously unreliable at this point. Is it still under
    > > warranty? Does Dell cover things like moisture incursion?
    > >
    > > If so I would be putting pressure on the to replace EVERYTHING. If it's
    > > no longer under warranty or covered - think real hard about getting a
    > > new one.
    > >
    > > That message indicates unrecoverable hard disk errors when data is read
    > > or written - you are losing data.
    > >
    > > I hope you are getting verified backups via Backup Edge or Lone-tar. I
    > > would not trust backups via tar or cpio.
    > >
    > > --
    > > ----------------------------------------------------
    > > Pat Welch, UBB Computer Services, a WCS Affiliate
    > > SCO Authorized Partner
    > > Unix/Linux/Windows/Hardware Sales/Support
    > > (209) 745-1401 Cell: (209) 251-9120
    > > E-mail: patubb@inreach.com
    > > ----------------------------------------------------

    >




  8. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk


    JP wrote:

    >
    > Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    > issues, just shut it down, pull every stinking board that is plugged in, dry
    > them by wiping, blowing, and in some instances using alcohol to make sure
    > that any and all visible water has been dried. Then use the old hair drier
    > to blow the port slots dry. Once this is done, let it sit for at least 24
    > hours with a fan blowing on it. Now you can re-insert the boards, boot it
    > up, make a good backup, shut it down, and replace the whole thing with a new
    > machine.
    >
    > The warning is this. WATER conducts electriciy and infiltrates in ways we
    > can not predict. Therefore, even by drying, we can not be sure that the
    > water has not caused unseen damage. When dealing with data integrity,
    > customer files, security related information, we can not and should not ever
    > assume things will be OK. We must KNOW they will. Thus, replacement of
    > equipment is the only reasonable answer.
    >
    > I could tell numerous horror stories of water damaged machines (I've been
    > involved with many), and none of them ends in story-book fashion.



    Ayup.

    Just last week I was helping someone by phone in adding a DVD drive to
    a system. We had gotten to the point where I had verified everything I
    needed to know before the physical install, so he said he'd call me
    back once it was installed.


    So he did.

    Because the machine was dirty when he opened it up, he decided to blow
    it out.
    Conveniently, this machine was located in an auto repair shop, which of
    course always has a nice supply of quite high pressure, oil and grease
    laden air. The high pressure is great for driving the oil and dirt
    deep into otherwise inaccessible places. Oil and grease are pretty
    good conductors, which is always helpful.

    Oh my. Error upon error, panic upon panic. That's what was going on
    when he called me back. I wasn't particularly surprised, expressed my
    condolences, and suggested new hardware.

    To my surprise, he called back several hours later, having somehow
    managed to get the machine working. I wouldn't ever trust the thing
    again, but we finished up the project and I left it at that..

    --
    Tony Lawrence
    http://aplawrence.com
    (Unix/Linux/Mac OSX resources etc.)


  9. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    Tony Lawrence wrote:
    > JP wrote:
    >
    >
    >>Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    >>issues, just shut it down, pull every stinking board that is plugged in, dry
    >>them by wiping, blowing, and in some instances using alcohol to make sure
    >>that any and all visible water has been dried. Then use the old hair drier
    >>to blow the port slots dry. Once this is done, let it sit for at least 24
    >>hours with a fan blowing on it. Now you can re-insert the boards, boot it
    >>up, make a good backup, shut it down, and replace the whole thing with a new
    >>machine.
    >>
    >>The warning is this. WATER conducts electriciy and infiltrates in ways we
    >>can not predict. Therefore, even by drying, we can not be sure that the
    >>water has not caused unseen damage. When dealing with data integrity,
    >>customer files, security related information, we can not and should not ever
    >>assume things will be OK. We must KNOW they will. Thus, replacement of
    >>equipment is the only reasonable answer.
    >>
    >>I could tell numerous horror stories of water damaged machines (I've been
    >>involved with many), and none of them ends in story-book fashion.

    >
    >
    >
    > Ayup.
    >
    > Just last week I was helping someone by phone in adding a DVD drive to
    > a system. We had gotten to the point where I had verified everything I
    > needed to know before the physical install, so he said he'd call me
    > back once it was installed.
    >
    >
    > So he did.
    >
    > Because the machine was dirty when he opened it up, he decided to blow
    > it out.
    > Conveniently, this machine was located in an auto repair shop, which of
    > course always has a nice supply of quite high pressure, oil and grease
    > laden air. The high pressure is great for driving the oil and dirt
    > deep into otherwise inaccessible places. Oil and grease are pretty
    > good conductors, which is always helpful.
    >
    > Oh my. Error upon error, panic upon panic. That's what was going on
    > when he called me back. I wasn't particularly surprised, expressed my
    > condolences, and suggested new hardware.
    >
    > To my surprise, he called back several hours later, having somehow
    > managed to get the machine working. I wouldn't ever trust the thing
    > again, but we finished up the project and I left it at that..
    >
    > --
    > Tony Lawrence
    > http://aplawrence.com
    > (Unix/Linux/Mac OSX resources etc.)
    >
    >
    >
    > .
    >

    In another life, I was installing a new video board into a computer at a
    pawnshop. The pawnshop manager brought out a leafblower(Gasoline engine
    powered) to blow the dust out of the computer. After rather hastily
    declining the use of the machine, already running, I just disconnected
    the computer and took it outside and shook out as much dust as possible
    and used a paper towel to pick up some of the rest of the dust, secured
    the cover and hooked it back up and left after it ran ok.

    --
    William P. Akers E-Mail: billa@mgmindustries.com
    MGM Industries, Inc. Web Site: http://www.mgmindustries.com
    Hendersonville TN

  10. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk


    billa wrote:
    > Tony Lawrence wrote:
    > > JP wrote:
    > >
    > >
    > >>Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    > >>issues, just shut it down, pull every stinking board that is plugged in, dry
    > >>them by wiping, blowing, and in some instances using alcohol to make sure
    > >>that any and all visible water has been dried. Then use the old hair drier
    > >>to blow the port slots dry. Once this is done, let it sit for at least 24
    > >>hours with a fan blowing on it. Now you can re-insert the boards, boot it
    > >>up, make a good backup, shut it down, and replace the whole thing with a new
    > >>machine.
    > >>
    > >>The warning is this. WATER conducts electriciy and infiltrates in ways we
    > >>can not predict. Therefore, even by drying, we can not be sure that the
    > >>water has not caused unseen damage. When dealing with data integrity,
    > >>customer files, security related information, we can not and should not ever
    > >>assume things will be OK. We must KNOW they will. Thus, replacement of
    > >>equipment is the only reasonable answer.
    > >>
    > >>I could tell numerous horror stories of water damaged machines (I've been
    > >>involved with many), and none of them ends in story-book fashion.

    > >
    > >
    > >
    > > Ayup.
    > >
    > > Just last week I was helping someone by phone in adding a DVD drive to
    > > a system. We had gotten to the point where I had verified everything I
    > > needed to know before the physical install, so he said he'd call me
    > > back once it was installed.
    > >
    > >
    > > So he did.
    > >
    > > Because the machine was dirty when he opened it up, he decided to blow
    > > it out.
    > > Conveniently, this machine was located in an auto repair shop, which of
    > > course always has a nice supply of quite high pressure, oil and grease
    > > laden air. The high pressure is great for driving the oil and dirt
    > > deep into otherwise inaccessible places. Oil and grease are pretty
    > > good conductors, which is always helpful.
    > >
    > > Oh my. Error upon error, panic upon panic. That's what was going on
    > > when he called me back. I wasn't particularly surprised, expressed my
    > > condolences, and suggested new hardware.
    > >
    > > To my surprise, he called back several hours later, having somehow
    > > managed to get the machine working. I wouldn't ever trust the thing
    > > again, but we finished up the project and I left it at that..
    > >
    > > --
    > > Tony Lawrence
    > > http://aplawrence.com
    > > (Unix/Linux/Mac OSX resources etc.)
    > >
    > >
    > >
    > > .
    > >

    > In another life, I was installing a new video board into a computer at a
    > pawnshop. The pawnshop manager brought out a leafblower(Gasoline engine
    > powered) to blow the dust out of the computer. After rather hastily
    > declining the use of the machine, already running, I just disconnected
    > the computer and took it outside and shook out as much dust as possible
    > and used a paper towel to pick up some of the rest of the dust, secured
    > the cover and hooked it back up and left after it ran ok.



    I guess the only thing worse would be blowing it out with a sand
    blaster... oh, wait, no I think they make something that blows tiny
    metal balls.. there, that out to clean 'er out!

    --
    Tony Lawrence
    http://aplawrence.com
    (Unix/Linux/Mac OSX resources etc.)


  11. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk


    ----- Original Message -----
    From: "Tony Lawrence"
    Newsgroups: comp.unix.sco.misc
    To:
    Sent: Monday, August 07, 2006 11:42 AM
    Subject: Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk


    >
    > JP wrote:
    >
    >>
    >> Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    >> issues, just shut it down, pull every stinking board that is plugged in,
    >> dry
    >> them by wiping, blowing, and in some instances using alcohol to make sure
    >> that any and all visible water has been dried. Then use the old hair
    >> drier
    >> to blow the port slots dry. Once this is done, let it sit for at least
    >> 24
    >> hours with a fan blowing on it. Now you can re-insert the boards, boot it
    >> up, make a good backup, shut it down, and replace the whole thing with a
    >> new
    >> machine.
    >>
    >> The warning is this. WATER conducts electriciy and infiltrates in ways
    >> we
    >> can not predict. Therefore, even by drying, we can not be sure that the
    >> water has not caused unseen damage. When dealing with data integrity,
    >> customer files, security related information, we can not and should not
    >> ever
    >> assume things will be OK. We must KNOW they will. Thus, replacement of
    >> equipment is the only reasonable answer.
    >>
    >> I could tell numerous horror stories of water damaged machines (I've been
    >> involved with many), and none of them ends in story-book fashion.

    >
    >
    > Ayup.
    >
    > Just last week I was helping someone by phone in adding a DVD drive to
    > a system. We had gotten to the point where I had verified everything I
    > needed to know before the physical install, so he said he'd call me
    > back once it was installed.
    >
    >
    > So he did.
    >
    > Because the machine was dirty when he opened it up, he decided to blow
    > it out.
    > Conveniently, this machine was located in an auto repair shop, which of
    > course always has a nice supply of quite high pressure, oil and grease
    > laden air. The high pressure is great for driving the oil and dirt
    > deep into otherwise inaccessible places. Oil and grease are pretty
    > good conductors, which is always helpful.
    >
    > Oh my. Error upon error, panic upon panic. That's what was going on
    > when he called me back. I wasn't particularly surprised, expressed my
    > condolences, and suggested new hardware.
    >
    > To my surprise, he called back several hours later, having somehow
    > managed to get the machine working. I wouldn't ever trust the thing
    > again, but we finished up the project and I left it at that..


    Oil and grease generally do not conduct electricity, and mechanic shops air
    compressors generally do not have oil and grease in the air.
    However water does conduct and all air compressors have water in their air.
    The act of compressing air squeases water out of it and so water collects
    everywhere that air gets compressed. This includes all along the air lines
    not just in the tank.
    There are "dryers" on the air lines to try to remove it but it's often not
    very effective. All the dryer is is a little tank to try to get air being
    compressed into the line to lose it's water there in the little tank before
    reaching the line. But really, every part of the system that has compressed
    air in it, also has air _being compressed_ in it, and wherever air gets
    compressed, water condenses out of it. So water collects all along the lines
    and blows out with the air.

    That also explains why it started working again after a while.

    Brian K. White -- brian@aljex.com -- http://www.aljex.com/bkw/
    +++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
    filePro BBx Linux SCO FreeBSD #callahans Satriani Filk!


  12. Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk

    In article <01b001c6ba57$e7b68f20$6500000a@venti>,
    Brian K. White wrote:
    >
    >----- Original Message -----
    >From: "Tony Lawrence"
    >Newsgroups: comp.unix.sco.misc
    >To:
    >Sent: Monday, August 07, 2006 11:42 AM
    >Subject: Re: NOTICE: Sdsk: Unrecoverable error writing SCSI disk
    >
    >
    >>
    >> JP wrote:
    >>
    >>>
    >>> Lesson learned I hope. When a machine, ANY MACHINE, has water integrity
    >>> issues, just shut it down, pull every stinking board that is plugged in,
    >>> dry
    >>> them by wiping, blowing, and in some instances using alcohol to make sure
    >>> that any and all visible water has been dried. Then use the old hair
    >>> drier
    >>> to blow the port slots dry. Once this is done, let it sit for at least
    >>> 24
    >>> hours with a fan blowing on it. Now you can re-insert the boards, boot it
    >>> up, make a good backup, shut it down, and replace the whole thing with a
    >>> new
    >>> machine.
    >>>
    >>> The warning is this. WATER conducts electriciy and infiltrates in ways
    >>> we
    >>> can not predict. Therefore, even by drying, we can not be sure that the
    >>> water has not caused unseen damage. When dealing with data integrity,
    >>> customer files, security related information, we can not and should not
    >>> ever
    >>> assume things will be OK. We must KNOW they will. Thus, replacement of
    >>> equipment is the only reasonable answer.
    >>>
    >>> I could tell numerous horror stories of water damaged machines (I've been
    >>> involved with many), and none of them ends in story-book fashion.

    >>
    >>
    >> Ayup.
    >>
    >> Just last week I was helping someone by phone in adding a DVD drive to
    >> a system. We had gotten to the point where I had verified everything I
    >> needed to know before the physical install, so he said he'd call me
    >> back once it was installed.
    >>
    >>
    >> So he did.
    >>
    >> Because the machine was dirty when he opened it up, he decided to blow
    >> it out.
    >> Conveniently, this machine was located in an auto repair shop, which of
    >> course always has a nice supply of quite high pressure, oil and grease
    >> laden air. The high pressure is great for driving the oil and dirt
    >> deep into otherwise inaccessible places. Oil and grease are pretty
    >> good conductors, which is always helpful.
    >>
    >> Oh my. Error upon error, panic upon panic. That's what was going on
    >> when he called me back. I wasn't particularly surprised, expressed my
    >> condolences, and suggested new hardware.
    >>
    >> To my surprise, he called back several hours later, having somehow
    >> managed to get the machine working. I wouldn't ever trust the thing
    >> again, but we finished up the project and I left it at that..

    >
    >Oil and grease generally do not conduct electricity, and mechanic shops air
    >compressors generally do not have oil and grease in the air.
    >However water does conduct and all air compressors have water in their air.
    >The act of compressing air squeases water out of it and so water collects
    >everywhere that air gets compressed. This includes all along the air lines
    >not just in the tank.
    >There are "dryers" on the air lines to try to remove it but it's often not
    >very effective. All the dryer is is a little tank to try to get air being
    >compressed into the line to lose it's water there in the little tank before
    >reaching the line. But really, every part of the system that has compressed
    >air in it, also has air _being compressed_ in it, and wherever air gets
    >compressed, water condenses out of it. So water collects all along the lines
    >and blows out with the air.
    >
    >That also explains why it started working again after a while.
    >
    >Brian K. White -- brian@aljex.com -- http://www.aljex.com/bkw/
    >+++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
    >filePro BBx Linux SCO FreeBSD #callahans Satriani Filk!


    And as the air expands it will cool and can cause condensation
    also.

    Bill


    --
    Bill Vermillion - bv @ wjv . com

+ Reply to Thread