SCSI disk medium error - SCO

This is a discussion on SCSI disk medium error - SCO ; on a 506 box (the one getting trap e errors, in fact) i see a few SCSI medium errors (slightly modified for clarity): .... NOTICE: Sdsk: Unrecoverable error reading SCSI disk 3 dev 1 /232 (ha=0 bus=0 id=3 lun=0) block=10623520 ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: SCSI disk medium error

  1. SCSI disk medium error

    on a 506 box (the one getting trap e errors, in fact)
    i see a few SCSI medium errors (slightly modified for
    clarity):

    .... NOTICE: Sdsk: Unrecoverable error reading SCSI disk 3 dev 1
    /232 (ha=0 bus=0 id=3 lun=0) block=10623520 ... Medium error

    seven such messages, covering nearby blocks. the application, rsync
    making an extra backup, blithely soldiered on, unaware of the
    problem. i assume that the SCSI subsystem corrected the errors, and
    that therefore the data is good, but I really don't know. so, i'd
    like to know which files resided on these seven blocks. now, i can
    tell which disk (and therefore which filesystem) was affected, but
    how can i tell which files are involved?

    i suppose i should run badtrk on this disk ASAP, and buy a new one.

    not a good day for hardware.

    thanks!

    --
    _________________________________________
    Nachman Yaakov Ziskind, FSPA, LLM awacs@ziskind.us
    Attorney and Counselor-at-Law http://ziskind.us
    Economic Group Pension Services http://egps.com
    Actuaries and Employee Benefit Consultants

  2. Re: SCSI disk medium error

    N. Yaakov Ziskind wrote:
    > on a 506 box (the one getting trap e errors, in fact)
    > i see a few SCSI medium errors (slightly modified for
    > clarity):
    >
    > ... NOTICE: Sdsk: Unrecoverable error reading SCSI disk 3 dev 1
    > /232 (ha=0 bus=0 id=3 lun=0) block=10623520 ... Medium error
    >
    > seven such messages, covering nearby blocks. the application, rsync
    > making an extra backup, blithely soldiered on, unaware of the
    > problem. i assume that the SCSI subsystem corrected the errors, and
    > that therefore the data is good, but I really don't know. so, i'd
    > like to know which files resided on these seven blocks. now, i can
    > tell which disk (and therefore which filesystem) was affected, but
    > how can i tell which files are involved?
    >
    > i suppose i should run badtrk on this disk ASAP, and buy a new one.


    I'd check the SCSI controller to see if it has a utility menu for
    "verifying" the hard disk. With Adaptec controllers, you press
    Ctl-A when prompted and enter the Easy SCSI setup. Adaptec Disk Utilities
    have format and verify. Verify the disk and it will spare out the bad
    sectors. Use badtrk only as the last resort. And be sure that the disk
    is unmounted (maintenance mode if the disk contains the root partition).

    Hopefully, your unnamed SCSI controller is an Adaptec or has the verify
    function.



    >
    > not a good day for hardware.
    >
    > thanks!
    >


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

  3. Re: SCSI disk medium error


    "N. Yaakov Ziskind" wrote in message
    news:20080618164302.B14051@egps.egps.com...
    > on a 506 box (the one getting trap e errors, in fact)
    > i see a few SCSI medium errors (slightly modified for
    > clarity):
    >
    > ... NOTICE: Sdsk: Unrecoverable error reading SCSI disk 3 dev 1
    > /232 (ha=0 bus=0 id=3 lun=0) block=10623520 ... Medium error
    >
    > seven such messages, covering nearby blocks. the application, rsync
    > making an extra backup, blithely soldiered on, unaware of the
    > problem. i assume that the SCSI subsystem corrected the errors, and
    > that therefore the data is good, but I really don't know. so, i'd
    > like to know which files resided on these seven blocks. now, i can
    > tell which disk (and therefore which filesystem) was affected, but
    > how can i tell which files are involved?
    >
    > i suppose i should run badtrk on this disk ASAP, and buy a new one.
    >
    > not a good day for hardware.
    >
    > thanks!


    It's unusual for many disparate devices to fail at once unless your
    power supply is failing. If you're using a no-name generic ps, try
    swapping in a known, preferably name-brand, new power supply.
    (I've always had a preference for PC Power & Cooling.)
    I assume this server is already connected to a UPS?

    Bob


  4. Re: SCSI disk medium error

    Bob Bailin wrote (on Wed, Jun 18, 2008 at 10:23:14PM -0400):
    >
    > "N. Yaakov Ziskind" wrote in message
    > news:20080618164302.B14051@egps.egps.com...
    > >on a 506 box (the one getting trap e errors, in fact)
    > >i see a few SCSI medium errors (slightly modified for
    > >clarity):
    > >
    > >... NOTICE: Sdsk: Unrecoverable error reading SCSI disk 3 dev 1
    > >/232 (ha=0 bus=0 id=3 lun=0) block=10623520 ... Medium error
    > >
    > >seven such messages, covering nearby blocks. the application, rsync
    > >making an extra backup, blithely soldiered on, unaware of the
    > >problem. i assume that the SCSI subsystem corrected the errors, and
    > >that therefore the data is good, but I really don't know. so, i'd
    > >like to know which files resided on these seven blocks. now, i can
    > >tell which disk (and therefore which filesystem) was affected, but
    > >how can i tell which files are involved?
    > >
    > >i suppose i should run badtrk on this disk ASAP, and buy a new one.
    > >
    > >not a good day for hardware.
    > >
    > >thanks!

    >
    > It's unusual for many disparate devices to fail at once unless your
    > power supply is failing. If you're using a no-name generic ps, try
    > swapping in a known, preferably name-brand, new power supply.
    > (I've always had a preference for PC Power & Cooling.)
    > I assume this server is already connected to a UPS?
    >
    > Bob


    True, but:
    a) server has a stock compaq ('more steel than some foreign cars') power
    supply;

    b) server resides on a UPS;

    c) the several crashes *could* have done damage to the disk.

    Right?

    --
    _________________________________________
    Nachman Yaakov Ziskind, FSPA, LLM awacs@ziskind.us
    Attorney and Counselor-at-Law http://ziskind.us
    Economic Group Pension Services http://egps.com
    Actuaries and Employee Benefit Consultants

  5. Re: SCSI disk medium error

    Steve M. Fabac, Jr. wrote (on Wed, Jun 18, 2008 at 05:21:06PM -0500):
    > N. Yaakov Ziskind wrote:
    > >on a 506 box (the one getting trap e errors, in fact)
    > >i see a few SCSI medium errors (slightly modified for
    > >clarity):
    > >
    > >... NOTICE: Sdsk: Unrecoverable error reading SCSI disk 3 dev 1
    > >/232 (ha=0 bus=0 id=3 lun=0) block=10623520 ... Medium error
    > >
    > >seven such messages, covering nearby blocks. the application, rsync
    > >making an extra backup, blithely soldiered on, unaware of the
    > >problem. i assume that the SCSI subsystem corrected the errors, and
    > >that therefore the data is good, but I really don't know. so, i'd
    > >like to know which files resided on these seven blocks. now, i can
    > >tell which disk (and therefore which filesystem) was affected, but
    > >how can i tell which files are involved?
    > >
    > >i suppose i should run badtrk on this disk ASAP, and buy a new one.

    >
    > I'd check the SCSI controller to see if it has a utility menu for
    > "verifying" the hard disk. With Adaptec controllers, you press
    > Ctl-A when prompted and enter the Easy SCSI setup. Adaptec Disk Utilities
    > have format and verify. Verify the disk and it will spare out the bad
    > sectors. Use badtrk only as the last resort. And be sure that the disk
    > is unmounted (maintenance mode if the disk contains the root partition).


    well, the attraction to badtrk is that you *can* run it while the server
    is running, as long as the disk is unmounted.

    it's not an adaptec, it's an embedded compaq, running the cha driver, so
    i'm sure compaq has diagnostic utilities; i'll crack them over the
    weekend.

    But back to my original question: does anyone have any idea how to
    figure out which files used the bad blocks in question? there *must* be
    a way to do it.

    > Hopefully, your unnamed SCSI controller is an Adaptec or has the verify
    > function.
    >
    > >not a good day for hardware.
    > >
    > >thanks!
    > >

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


    --
    _________________________________________
    Nachman Yaakov Ziskind, FSPA, LLM awacs@ziskind.us
    Attorney and Counselor-at-Law http://ziskind.us
    Economic Group Pension Services http://egps.com
    Actuaries and Employee Benefit Consultants

  6. Re: SCSI disk medium error

    N. Yaakov Ziskind wrote:

    > But back to my original question: does anyone have any idea how to
    > figure out which files used the bad blocks in question? there *must* be
    > a way to do it.


    badblk(ADM), but I'm not sure what state of repair it's in...

    >Bela<


+ Reply to Thread