Need Suggestons on SCSI drives "not ready" - SCO

This is a discussion on Need Suggestons on SCSI drives "not ready" - SCO ; I have a client running SCO 5.0.5 with DPT 3754U2 RAID controller with two disks in RAID1. The problem occurred when the client's building lost power and the UPS battery expired after it ran down. No, there was no signaling ...

+ Reply to Thread
Results 1 to 11 of 11

Thread: Need Suggestons on SCSI drives "not ready"

  1. Need Suggestons on SCSI drives "not ready"

    I have a client running SCO 5.0.5 with DPT 3754U2 RAID
    controller with two disks in RAID1.

    The problem occurred when the client's building lost power and
    the UPS battery expired after it ran down. No, there was
    no signaling cable to the SCO box and therefore no way to
    detect the power failure. (This will be rectified).

    But the problem is the two Seagate 140G 15K cheetah drives.
    They are down hard and won't come up. These are brand new
    drives and have only been in the machine for a week.

    I pulled the drives and replaced them with a single drive on
    the DPT controller, booted the Microlite RE2 media and restored
    the system from the previous night's nightly backup tape.

    Seagate says get an RMA and they will replace the drives. However,
    I would like to get them working so that I can grab the log files
    and application data files (one days worth of orders and shipping,
    since the last backup, not enough value to send the drives to a
    data recovery service).

    Also, I would like to get the drives back up for the knowledge on
    how to do it in the future.

    The problem is when you boot the system with the failed drives and
    press Ctl-D to enter the DPT RAID configuration utility, the drives
    are seen but are listed with the block size field showing "no media."
    (This normally shows 512 bytes). THE DPT controller lists the drives
    as "optimal" and not "failed." Also, the SCSI activity light on the
    drive carrier is "on" solid and never goes out.

    When I connected the SCA cage to the on-board Adpatec U160 controller
    and booted the system, the U160 splash screen shows the drives on
    their respective ID's but then adds: "Failed Start Unit Request."
    Pressing Ctl-A to enter Adaptec Easy SCSI configuration screen and
    selecting SCSI utilities shows the drives but trying to verify the
    drive results in:

    SCSI CDB sent: 03 00 00 0E 00 70 00 02 00
    SCSI Controller Status: 00h No SCSI Controller Error
    Target Status: 02h Check condition
    Sense Key: 02h Not Ready
    Sense Code: 04h
    Sense Code Qualifier: 00h


    The fdisk layout has a 300 track DOS partition from track 1 to track 300
    with the active SCO boot partition starting at track 301. If this were
    an IDE disk, I would stomp on the first 10 or so tracks, and then recreate
    the fdisk partition layout and then the Divvy tables without creating
    new file systems. However, since the disk is SCSI and won't respond to the
    "start unit request" I can't read or write to the disk with any tools I
    have: Booting a DOS floppy with config.sys containing device=aspi8u2.sys
    (device driver for Ultra2 and Ultra160 class host adapters) does not
    provide access to the disk.

    I'm hopeful that someone on the group has experience with this problem and
    can point me to recovery tools that I can get to wake these disks back up.

    And FYI: Calling Adaptec technical support on the DPT 3754U2 controllers
    was a bust. The customer service agent indicated that "Adaptec does no provide
    technical support on obsolete products. You may be able to find some help in the
    Adaptec ASK database."

    Seagate technical support advised me to download the Seagate SeaTools boot ISO
    image but that again only talks to IDE drives. They have a SeaTools Enterprise
    edition that runs under Windows or Linux but that requires that your SCSI
    disk and controller be installed in a Windows or Linux system. Not an option
    for me.


    --
    Steve Fabac
    S.M. Fabac & Associates
    816/765-1670

  2. Re: Need Suggestons on SCSI drives "not ready"

    Steve M. Fabac, Jr. wrote:

    > I'm hopeful that someone on the group has experience with this problem and
    > can point me to recovery tools that I can get to wake these disks back up.
    >
    > And FYI: Calling Adaptec technical support on the DPT 3754U2 controllers
    > was a bust. The customer service agent indicated that "Adaptec does no
    > provide
    > technical support on obsolete products. You may be able to find some
    > help in the
    > Adaptec ASK database."
    >
    > Seagate technical support advised me to download the Seagate SeaTools
    > boot ISO
    > image but that again only talks to IDE drives. They have a SeaTools
    > Enterprise
    > edition that runs under Windows or Linux but that requires that your SCSI
    > disk and controller be installed in a Windows or Linux system. Not an
    > option
    > for me.
    >


    Is it possible to boot a Knoppix "Live" cd or other similar Linux
    bootable cd so you can run the Seagate Enterprise app on the actual SCO
    server? If it needs real disk space to install on just use some form of
    USB drive.

    Good Luck

    Glenn

  3. Re: Need Suggestons on SCSI drives "not ready"

    Glenn wrote:
    > Steve M. Fabac, Jr. wrote:
    >
    >> I'm hopeful that someone on the group has experience with this problem
    >> and
    >> can point me to recovery tools that I can get to wake these disks back
    >> up.
    >>
    >> And FYI: Calling Adaptec technical support on the DPT 3754U2 controllers
    >> was a bust. The customer service agent indicated that "Adaptec does no
    >> provide
    >> technical support on obsolete products. You may be able to find some
    >> help in the
    >> Adaptec ASK database."
    >>
    >> Seagate technical support advised me to download the Seagate SeaTools
    >> boot ISO
    >> image but that again only talks to IDE drives. They have a SeaTools
    >> Enterprise
    >> edition that runs under Windows or Linux but that requires that your SCSI
    >> disk and controller be installed in a Windows or Linux system. Not an
    >> option
    >> for me.
    >>

    >
    > Is it possible to boot a Knoppix "Live" cd or other similar Linux
    > bootable cd so you can run the Seagate Enterprise app on the actual SCO
    > server? If it needs real disk space to install on just use some form of
    > USB drive.


    Good suggestion. I moved the SCA cage back to the Adapted 29160 controller
    and booted DSL (Damn Small Linux) CD and used firefox to download the SeaTools
    enterprise for Linux version and ran it.

    The "st -I" shows all the devices on the SCSI bus and identifies the disk
    correctly. However, It says "can't get capacity" and "sense error 0/2/0"

    "st -t /dev/sgb" command fails with the message "...does not support
    DST" Drive Self Test.

    I also tried the st -F to low level format one of the drives but that
    failed with repeated lines of "formatting ..... 0% completed" and repeated
    over and over until I aborted.

    The last Seagate enterprise support technician suggested to use the
    "zero fill" option to write zeros to the first several sectors but
    I was unable to find the zero fill option.

    So far no success.


    >
    > Good Luck
    >
    > Glenn
    >
    >


    --
    Steve Fabac
    S.M. Fabac & Associates
    816/765-1670

  4. Re: Need Suggestons on SCSI drives "not ready"

    Steve M. Fabac, Jr. wrote:
    > Glenn wrote:
    >> Steve M. Fabac, Jr. wrote:
    >>
    >>> I'm hopeful that someone on the group has experience with this
    >>> problem and
    >>> can point me to recovery tools that I can get to wake these disks
    >>> back up.
    >>>
    >>> And FYI: Calling Adaptec technical support on the DPT 3754U2 controllers
    >>> was a bust. The customer service agent indicated that "Adaptec does
    >>> no provide
    >>> technical support on obsolete products. You may be able to find some
    >>> help in the
    >>> Adaptec ASK database."
    >>>
    >>> Seagate technical support advised me to download the Seagate SeaTools
    >>> boot ISO
    >>> image but that again only talks to IDE drives. They have a SeaTools
    >>> Enterprise
    >>> edition that runs under Windows or Linux but that requires that your
    >>> SCSI
    >>> disk and controller be installed in a Windows or Linux system. Not an
    >>> option
    >>> for me.
    >>>

    >>
    >> Is it possible to boot a Knoppix "Live" cd or other similar Linux
    >> bootable cd so you can run the Seagate Enterprise app on the actual
    >> SCO server? If it needs real disk space to install on just use some
    >> form of USB drive.

    >
    > Good suggestion. I moved the SCA cage back to the Adapted 29160 controller
    > and booted DSL (Damn Small Linux) CD and used firefox to download the
    > SeaTools
    > enterprise for Linux version and ran it.
    >
    > The "st -I" shows all the devices on the SCSI bus and identifies the disk
    > correctly. However, It says "can't get capacity" and "sense error 0/2/0"
    >
    > "st -t /dev/sgb" command fails with the message "...does not support
    > DST" Drive Self Test.
    >
    > I also tried the st -F to low level format one of the drives but that
    > failed with repeated lines of "formatting ..... 0% completed" and repeated
    > over and over until I aborted.
    >
    > The last Seagate enterprise support technician suggested to use the
    > "zero fill" option to write zeros to the first several sectors but
    > I was unable to find the zero fill option.


    Would 'dd if=/dev/zero of=/dev/sda' or the appropriate SCSI devices, do the
    job for you? And do you need to recover *any* of this data? A data recovery
    house will have some fascinating, but expensive, tools to do just this.

  5. Re: Need Suggestons on SCSI drives "not ready"


    "Steve M. Fabac, Jr." wrote in message
    news:485D63FB.4030803@att.net...
    >I have a client running SCO 5.0.5 with DPT 3754U2 RAID
    > controller with two disks in RAID1.
    >
    > The problem occurred when the client's building lost power and
    > the UPS battery expired after it ran down. No, there was
    > no signaling cable to the SCO box and therefore no way to
    > detect the power failure. (This will be rectified).
    >
    > But the problem is the two Seagate 140G 15K cheetah drives.
    > They are down hard and won't come up. These are brand new
    > drives and have only been in the machine for a week.
    >
    > I pulled the drives and replaced them with a single drive on
    > the DPT controller, booted the Microlite RE2 media and restored
    > the system from the previous night's nightly backup tape.
    >
    > Seagate says get an RMA and they will replace the drives. However,
    > I would like to get them working so that I can grab the log files
    > and application data files (one days worth of orders and shipping,
    > since the last backup, not enough value to send the drives to a
    > data recovery service).
    >
    > Also, I would like to get the drives back up for the knowledge on
    > how to do it in the future.
    >
    > The problem is when you boot the system with the failed drives and
    > press Ctl-D to enter the DPT RAID configuration utility, the drives
    > are seen but are listed with the block size field showing "no media."
    > (This normally shows 512 bytes). THE DPT controller lists the drives
    > as "optimal" and not "failed." Also, the SCSI activity light on the
    > drive carrier is "on" solid and never goes out.
    >
    > When I connected the SCA cage to the on-board Adpatec U160 controller
    > and booted the system, the U160 splash screen shows the drives on
    > their respective ID's but then adds: "Failed Start Unit Request."
    > Pressing Ctl-A to enter Adaptec Easy SCSI configuration screen and
    > selecting SCSI utilities shows the drives but trying to verify the
    > drive results in:
    >
    > SCSI CDB sent: 03 00 00 0E 00 70 00 02 00
    > SCSI Controller Status: 00h No SCSI Controller Error
    > Target Status: 02h Check condition
    > Sense Key: 02h Not Ready
    > Sense Code: 04h
    > Sense Code Qualifier: 00h
    >
    >
    > The fdisk layout has a 300 track DOS partition from track 1 to track 300
    > with the active SCO boot partition starting at track 301. If this were
    > an IDE disk, I would stomp on the first 10 or so tracks, and then recreate
    > the fdisk partition layout and then the Divvy tables without creating
    > new file systems. However, since the disk is SCSI and won't respond to the
    > "start unit request" I can't read or write to the disk with any tools I
    > have: Booting a DOS floppy with config.sys containing device=aspi8u2.sys
    > (device driver for Ultra2 and Ultra160 class host adapters) does not
    > provide access to the disk.
    >
    > I'm hopeful that someone on the group has experience with this problem and
    > can point me to recovery tools that I can get to wake these disks back up.
    >
    > And FYI: Calling Adaptec technical support on the DPT 3754U2 controllers
    > was a bust. The customer service agent indicated that "Adaptec does no
    > provide
    > technical support on obsolete products. You may be able to find some help
    > in the
    > Adaptec ASK database."
    >
    > Seagate technical support advised me to download the Seagate SeaTools boot
    > ISO
    > image but that again only talks to IDE drives. They have a SeaTools
    > Enterprise
    > edition that runs under Windows or Linux but that requires that your SCSI
    > disk and controller be installed in a Windows or Linux system. Not an
    > option
    > for me.


    Have you tried changing the jumpers on the drive so that they don't wait for
    a "Start Unit" command? Without this delay (usually a multiple of the
    drive's
    SCSI ID) the drive should spin up upon receiving power.

    While you're at it, check the other jumpers to ensure that no other shunts
    have
    accidently fallen off.

    Bob


  6. Re: Need Suggestons on SCSI drives "not ready"

    Nico Kadel-Garcia wrote:
    > Steve M. Fabac, Jr. wrote:
    >> Glenn wrote:
    >>> Steve M. Fabac, Jr. wrote:
    >>>
    >>>> I'm hopeful that someone on the group has experience with this
    >>>> problem and
    >>>> can point me to recovery tools that I can get to wake these disks
    >>>> back up.
    >>>>
    >>>> And FYI: Calling Adaptec technical support on the DPT 3754U2
    >>>> controllers
    >>>> was a bust. The customer service agent indicated that "Adaptec does
    >>>> no provide
    >>>> technical support on obsolete products. You may be able to find some
    >>>> help in the
    >>>> Adaptec ASK database."
    >>>>
    >>>> Seagate technical support advised me to download the Seagate
    >>>> SeaTools boot ISO
    >>>> image but that again only talks to IDE drives. They have a SeaTools
    >>>> Enterprise
    >>>> edition that runs under Windows or Linux but that requires that your
    >>>> SCSI
    >>>> disk and controller be installed in a Windows or Linux system. Not
    >>>> an option
    >>>> for me.
    >>>>
    >>>
    >>> Is it possible to boot a Knoppix "Live" cd or other similar Linux
    >>> bootable cd so you can run the Seagate Enterprise app on the actual
    >>> SCO server? If it needs real disk space to install on just use some
    >>> form of USB drive.

    >>
    >> Good suggestion. I moved the SCA cage back to the Adapted 29160
    >> controller
    >> and booted DSL (Damn Small Linux) CD and used firefox to download the
    >> SeaTools
    >> enterprise for Linux version and ran it.
    >>
    >> The "st -I" shows all the devices on the SCSI bus and identifies the disk
    >> correctly. However, It says "can't get capacity" and "sense error 0/2/0"
    >>
    >> "st -t /dev/sgb" command fails with the message "...does not support
    >> DST" Drive Self Test.
    >>
    >> I also tried the st -F to low level format one of the drives but that
    >> failed with repeated lines of "formatting ..... 0% completed" and
    >> repeated
    >> over and over until I aborted.
    >>
    >> The last Seagate enterprise support technician suggested to use the
    >> "zero fill" option to write zeros to the first several sectors but
    >> I was unable to find the zero fill option.

    >
    > Would 'dd if=/dev/zero of=/dev/sda' or the appropriate SCSI devices, do
    > the job for you? And do you need to recover *any* of this data? A data


    root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    1+0 records in
    1+0 records out
    root@1[root]#
    root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=100
    dd: /dev/sga: Numerical argument out of domain
    root@1[root]#
    root@1[root]# fuser /dev/sga
    /dev/sga: 581
    root@1[root]# ps -ef | grep 581
    581 root 512 S st -F /dev/sga
    root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    1+0 records in
    1+0 records out
    root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=2
    dd: /dev/sga: Numerical argument out of domain
    root@1[root]#

    st -I produces:

    /dev/sga [=/dev/sda bus1 ch=0 target=0 lun=0]
    /dev/sga
    Vendor = SEAGATE
    Product = ST3146855LC
    Version = 0003
    Serial Number = 3LN5044Z
    Copyright = Copyright (c) 2006 Seagate All rights reserved
    SCSI Firmware = 12200003
    Servo RAM Release = 2007C591
    Servo ROM Release = 00000000
    Servo RAM Date = C591
    Servo ROM Date = 2007
    -this is a Seagate drive
    -this drive does not support DST
    -Mode Page Settings [current value (default)]:
    -WCE bit = 1 (0)
    -RCD bit = 1 (0)
    -AWRE bit = 0 (0)
    -ARRE bit = 1 (0)
    -DExcpt bit = 0 (0)
    -Number of cache segments = 76 (-117)
    -JIT bit 0 = 1 (1)
    -JIT bit 1 = 1 (1)
    -JIT bit 2 = 0 (1)
    -JIT bit 3 = 0 (1)

    and root@0[root]# st -i /dev/sga | less produces:
    /dev/sga
    Vendor = SEAGATE
    Product = ST3146855LC
    Version = 0003
    Serial Number = 3LN5044Z
    Copyright = Copyright (c) 2006 Seagate All rights reserved
    SCSI Firmware = 12200003
    Servo RAM Release = 2007C591
    Servo ROM Release = 00000000
    Servo RAM Date = C591
    Servo ROM Date = 2007
    > -Cannot read capacity (Sense data = 02/04/00)

    -this is a Seagate drive
    -this drive does not support DST
    -Mode Page Settings [current value (default)]:
    -WCE bit = 0 (1)
    -RCD bit = 1 (0)
    -AWRE bit = 0 (0)
    -ARRE bit = 1 (0)
    -DExcpt bit = 1 (0)
    -Number of cache segments = 116 (0)
    -JIT bit 1 = 1 (1)
    -JIT bit 2 = 0 (1)
    -JIT bit 3 = 0 (1)
    (END)


    root@0[root]# st -g 10% /dev/sga
    Starting 10 % Generic Short Test on drive /dev/sga (^C will abort test)
    test FAILED - sense data = 02/04/00
    Generic Short Test FAILED on drive /dev/sga

    root@0[root]# st -C max /dev/sga
    Setting drive /dev/sga highblock to max block
    root@0[root]# st -B 512 /dev/sga

    WARNING!!! Changing blocksize will ERASE all data on drive /dev/sga
    Enter Y to change blocksize to 512, any other key to abort - Y
    Setting drive /dev/sga to blocksize 512 bytes/block
    root@0[root]#

    Trying a desperation move and format one of the RAID-1 disks, Each line
    below at 0% is repeated every 60 seconds. I see no evidence on the drive
    that a format is occurring. Normally a format produces a dim glow on the
    the activity led. This drive is solid on without change.

    st -F /dev/sga
    FORMAT WILL DESTROY ALL DATA on drive /dev/sga
    - type ''Y '' to format, any other key to abort - Y
    Are you REALLY sure you want to format drive /dev/sga? - Y


    Starting FORMAT of drive /dev/sga . . . DO NOT INTERRUPT!!!
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .
    FORMAT on drive /dev/sga is 0 % complete . . .


    I suppose I'll let it run all night to see if there is any change
    in the drives status.


    > recovery house will have some fascinating, but expensive, tools to do
    > just this.


    No, a full backup is preformed nightly at 03:00. The loss in restoring
    the nightly backup is just today's orders and shipping and invoicing
    details. Once I restored the backup to a new disk and turned the machine
    back over to the client, the office and plant workers went to work re-entering
    the days transactions from hard copy.

    The only thing left to do is to get these drives back up if I can, or
    get an RMA from Seagate and replace them.

    I'm spending time trying to get the drives back up for the knowledge
    to be gained if I succeed and thus possibly armed with the tools to
    do it in future situations where the data is not backed up and needs to
    be recovered.
    >
    >


    --
    Steve Fabac
    S.M. Fabac & Associates
    816/765-1670


  7. Re: Need Suggestons on SCSI drives "not ready"

    Steve M. Fabac, Jr. wrote:
    > Nico Kadel-Garcia wrote:
    >> Steve M. Fabac, Jr. wrote:
    >>> Glenn wrote:
    >>>> Steve M. Fabac, Jr. wrote:
    >>>>
    >>>>> I'm hopeful that someone on the group has experience with this
    >>>>> problem and
    >>>>> can point me to recovery tools that I can get to wake these disks
    >>>>> back up.
    >>>>>
    >>>>> And FYI: Calling Adaptec technical support on the DPT 3754U2
    >>>>> controllers
    >>>>> was a bust. The customer service agent indicated that "Adaptec does
    >>>>> no provide
    >>>>> technical support on obsolete products. You may be able to find
    >>>>> some help in the
    >>>>> Adaptec ASK database."
    >>>>>
    >>>>> Seagate technical support advised me to download the Seagate
    >>>>> SeaTools boot ISO
    >>>>> image but that again only talks to IDE drives. They have a SeaTools
    >>>>> Enterprise
    >>>>> edition that runs under Windows or Linux but that requires that
    >>>>> your SCSI
    >>>>> disk and controller be installed in a Windows or Linux system. Not
    >>>>> an option
    >>>>> for me.
    >>>>>
    >>>>
    >>>> Is it possible to boot a Knoppix "Live" cd or other similar Linux
    >>>> bootable cd so you can run the Seagate Enterprise app on the actual
    >>>> SCO server? If it needs real disk space to install on just use some
    >>>> form of USB drive.
    >>>
    >>> Good suggestion. I moved the SCA cage back to the Adapted 29160
    >>> controller
    >>> and booted DSL (Damn Small Linux) CD and used firefox to download
    >>> the SeaTools
    >>> enterprise for Linux version and ran it.
    >>>
    >>> The "st -I" shows all the devices on the SCSI bus and identifies the
    >>> disk
    >>> correctly. However, It says "can't get capacity" and "sense error 0/2/0"
    >>>
    >>> "st -t /dev/sgb" command fails with the message "...does not support
    >>> DST" Drive Self Test.
    >>>
    >>> I also tried the st -F to low level format one of the drives but that
    >>> failed with repeated lines of "formatting ..... 0% completed" and
    >>> repeated
    >>> over and over until I aborted.
    >>>
    >>> The last Seagate enterprise support technician suggested to use the
    >>> "zero fill" option to write zeros to the first several sectors but
    >>> I was unable to find the zero fill option.

    >>
    >> Would 'dd if=/dev/zero of=/dev/sda' or the appropriate SCSI devices,
    >> do the job for you? And do you need to recover *any* of this data? A data

    >
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    > 1+0 records in
    > 1+0 records out
    > root@1[root]#
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=100
    > dd: /dev/sga: Numerical argument out of domain
    > root@1[root]#
    > root@1[root]# fuser /dev/sga
    > /dev/sga: 581
    > root@1[root]# ps -ef | grep 581
    > 581 root 512 S st -F /dev/sga
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    > 1+0 records in
    > 1+0 records out
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=2
    > dd: /dev/sga: Numerical argument out of domain
    > root@1[root]#


    Why not simply say 'dd if=/dev/zero of=/dev/sga'? You want to zero the whole
    disk if possible, and if you have the time.

  8. Re: Need Suggestons on SCSI drives "not ready"

    My only similar issue used the screwdriver blade to drive , handle to
    ear, and the lack of whine as proof the drive was not spinning up. Got
    access via adjacent drive mounting space.

  9. Re: Need Suggestons on SCSI drives "not ready"

    Steve M. Fabac, Jr. wrote:
    > Nico Kadel-Garcia wrote:
    >> Steve M. Fabac, Jr. wrote:
    >>> Glenn wrote:
    >>>> Steve M. Fabac, Jr. wrote:
    >>>>
    >>>>> I'm hopeful that someone on the group has experience with this
    >>>>> problem and
    >>>>> can point me to recovery tools that I can get to wake these disks
    >>>>> back up.

    ....
    >>> The last Seagate enterprise support technician suggested to use the
    >>> "zero fill" option to write zeros to the first several sectors but
    >>> I was unable to find the zero fill option.

    >>
    >> Would 'dd if=/dev/zero of=/dev/sda' or the appropriate SCSI devices,
    >> do the job for you? And do you need to recover *any* of this data? A data

    >
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    > 1+0 records in
    > 1+0 records out
    > root@1[root]#
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=100
    > dd: /dev/sga: Numerical argument out of domain
    > root@1[root]#
    > root@1[root]# fuser /dev/sga
    > /dev/sga: 581
    > root@1[root]# ps -ef | grep 581
    > 581 root 512 S st -F /dev/sga
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    > 1+0 records in
    > 1+0 records out
    > root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=2
    > dd: /dev/sga: Numerical argument out of domain
    > root@1[root]#
    >
    > st -I produces:
    >
    > /dev/sga [=/dev/sda bus1 ch=0 target=0 lun=0]
    > /dev/sga
    > Vendor = SEAGATE
    > Product = ST3146855LC
    > Version = 0003
    > Serial Number = 3LN5044Z
    > Copyright = Copyright (c) 2006 Seagate All rights reserved
    > SCSI Firmware = 12200003
    > Servo RAM Release = 2007C591
    > Servo ROM Release = 00000000
    > Servo RAM Date = C591
    > Servo ROM Date = 2007
    > -this is a Seagate drive
    > -this drive does not support DST
    > -Mode Page Settings [current value (default)]:
    > -WCE bit = 1 (0)
    > -RCD bit = 1 (0)
    > -AWRE bit = 0 (0)
    > -ARRE bit = 1 (0)
    > -DExcpt bit = 0 (0)
    > -Number of cache segments = 76 (-117)
    > -JIT bit 0 = 1 (1)
    > -JIT bit 1 = 1 (1)
    > -JIT bit 2 = 0 (1)
    > -JIT bit 3 = 0 (1)
    >
    > and root@0[root]# st -i /dev/sga | less produces:
    > /dev/sga
    > Vendor = SEAGATE
    > Product = ST3146855LC
    > Version = 0003
    > Serial Number = 3LN5044Z
    > Copyright = Copyright (c) 2006 Seagate All rights reserved
    > SCSI Firmware = 12200003
    > Servo RAM Release = 2007C591
    > Servo ROM Release = 00000000
    > Servo RAM Date = C591
    > Servo ROM Date = 2007
    >> -Cannot read capacity (Sense data = 02/04/00)

    > -this is a Seagate drive
    > -this drive does not support DST
    > -Mode Page Settings [current value (default)]:
    > -WCE bit = 0 (1)
    > -RCD bit = 1 (0)
    > -AWRE bit = 0 (0)
    > -ARRE bit = 1 (0)
    > -DExcpt bit = 1 (0)
    > -Number of cache segments = 116 (0)
    > -JIT bit 1 = 1 (1)
    > -JIT bit 2 = 0 (1)
    > -JIT bit 3 = 0 (1)
    > (END)
    >
    >
    > root@0[root]# st -g 10% /dev/sga
    > Starting 10 % Generic Short Test on drive /dev/sga (^C will abort test)
    > test FAILED - sense data = 02/04/00
    > Generic Short Test FAILED on drive /dev/sga
    >
    > root@0[root]# st -C max /dev/sga
    > Setting drive /dev/sga highblock to max block
    > root@0[root]# st -B 512 /dev/sga
    >
    > WARNING!!! Changing blocksize will ERASE all data on drive /dev/sga
    > Enter Y to change blocksize to 512, any other key to abort - Y
    > Setting drive /dev/sga to blocksize 512 bytes/block
    > root@0[root]#
    >
    > Trying a desperation move and format one of the RAID-1 disks, Each line
    > below at 0% is repeated every 60 seconds. I see no evidence on the drive
    > that a format is occurring. Normally a format produces a dim glow on the
    > the activity led. This drive is solid on without change.
    >
    > st -F /dev/sga
    > FORMAT WILL DESTROY ALL DATA on drive /dev/sga
    > - type ''Y '' to format, any other key to abort - Y
    > Are you REALLY sure you want to format drive /dev/sga? - Y
    >
    >
    > Starting FORMAT of drive /dev/sga . . . DO NOT INTERRUPT!!!
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    > FORMAT on drive /dev/sga is 0 % complete . . .
    >
    >
    > I suppose I'll let it run all night to see if there is any change
    > in the drives status.


    After 8 hours:

    FORMAT on drive /dev/sga is 0 % complete . . .
    read_capacity (SG_IO) error: Bad file descriptor
    FORMAT on drive /dev/sga is 0 % complete . . .
    read_capacity (SG_IO) error: Bad file descriptor
    FORMAT on drive /dev/sga is 0 % complete . . .
    read_capacity (SG_IO) error: Bad file descriptor
    FORMAT on drive /dev/sga is 0 % complete . . .
    read_capacity (SG_IO) error: Bad file descriptor
    FORMAT on drive /dev/sga is 0 % complete . . .
    read_capacity (SG_IO) error: Bad file descriptor
    FORMAT on drive /dev/sga is 0 % complete . . .



    --
    Steve Fabac
    S.M. Fabac & Associates
    816/765-1670

  10. Re: Need Suggestons on SCSI drives "not ready"

    Nico Kadel-Garcia wrote:
    > Steve M. Fabac, Jr. wrote:
    >> Nico Kadel-Garcia wrote:
    >>> Steve M. Fabac, Jr. wrote:
    >>>> Glenn wrote:
    >>>>> Steve M. Fabac, Jr. wrote:
    >>>>>
    >>>>>> I'm hopeful that someone on the group has experience with this
    >>>>>> problem and
    >>>>>> can point me to recovery tools that I can get to wake these disks
    >>>>>> back up.
    >>>>>>
    >>>>>> And FYI: Calling Adaptec technical support on the DPT 3754U2
    >>>>>> controllers
    >>>>>> was a bust. The customer service agent indicated that "Adaptec
    >>>>>> does no provide
    >>>>>> technical support on obsolete products. You may be able to find
    >>>>>> some help in the
    >>>>>> Adaptec ASK database."
    >>>>>>
    >>>>>> Seagate technical support advised me to download the Seagate
    >>>>>> SeaTools boot ISO
    >>>>>> image but that again only talks to IDE drives. They have a
    >>>>>> SeaTools Enterprise
    >>>>>> edition that runs under Windows or Linux but that requires that
    >>>>>> your SCSI
    >>>>>> disk and controller be installed in a Windows or Linux system. Not
    >>>>>> an option
    >>>>>> for me.
    >>>>>>
    >>>>>
    >>>>> Is it possible to boot a Knoppix "Live" cd or other similar Linux
    >>>>> bootable cd so you can run the Seagate Enterprise app on the
    >>>>> actual SCO server? If it needs real disk space to install on just
    >>>>> use some form of USB drive.
    >>>>
    >>>> Good suggestion. I moved the SCA cage back to the Adapted 29160
    >>>> controller
    >>>> and booted DSL (Damn Small Linux) CD and used firefox to download
    >>>> the SeaTools
    >>>> enterprise for Linux version and ran it.
    >>>>
    >>>> The "st -I" shows all the devices on the SCSI bus and identifies the
    >>>> disk
    >>>> correctly. However, It says "can't get capacity" and "sense error
    >>>> 0/2/0"
    >>>>
    >>>> "st -t /dev/sgb" command fails with the message "...does not support
    >>>> DST" Drive Self Test.
    >>>>
    >>>> I also tried the st -F to low level format one of the drives but that
    >>>> failed with repeated lines of "formatting ..... 0% completed" and
    >>>> repeated
    >>>> over and over until I aborted.
    >>>>
    >>>> The last Seagate enterprise support technician suggested to use the
    >>>> "zero fill" option to write zeros to the first several sectors but
    >>>> I was unable to find the zero fill option.
    >>>
    >>> Would 'dd if=/dev/zero of=/dev/sda' or the appropriate SCSI devices,
    >>> do the job for you? And do you need to recover *any* of this data? A
    >>> data

    >>
    >> root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    >> 1+0 records in
    >> 1+0 records out
    >> root@1[root]#
    >> root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=100
    >> dd: /dev/sga: Numerical argument out of domain
    >> root@1[root]#
    >> root@1[root]# fuser /dev/sga
    >> /dev/sga: 581
    >> root@1[root]# ps -ef | grep 581
    >> 581 root 512 S st -F /dev/sga
    >> root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=1
    >> 1+0 records in
    >> 1+0 records out
    >> root@1[root]# dd if=/dev/zero of=/dev/sga bs=512 count=2
    >> dd: /dev/sga: Numerical argument out of domain
    >> root@1[root]#

    >
    > Why not simply say 'dd if=/dev/zero of=/dev/sga'? You want to zero the
    > whole disk if possible, and if you have the time.
    >


    When I do, I get the following:

    root@0[root]# dd if=/dev/zero of=/dev/sga
    dd: /dev/sga: Numerical argument out of domain
    root@0[root]#

    What does the "Numerical argument out of domain" tell you?


    --
    Steve Fabac
    S.M. Fabac & Associates
    816/765-1670

  11. Re: Need Suggestons on SCSI drives "not ready"

    Steve M. Fabac, Jr. wrote:
    > But the problem is the two Seagate 140G 15K cheetah drives.
    > They are down hard and won't come up. These are brand new
    > drives and have only been in the machine for a week.


    Are the drives spinning or not? If not, you try with a gentle shock, or
    opening them and see if you can manually kick start the mechanism...

+ Reply to Thread