FIle corruption on HPFS partition - OS2

This is a discussion on FIle corruption on HPFS partition - OS2 ; On a FP15 system, I have an HPFS partition containing numerous small files (java source files). These files are frequently edited and compiled, so there is a lot of file activity on this partition. Recently I have started getting situations ...

+ Reply to Thread
Results 1 to 10 of 10

Thread: FIle corruption on HPFS partition

  1. FIle corruption on HPFS partition

    On a FP15 system, I have an HPFS partition containing numerous small
    files (java source files). These files are frequently edited and
    compiled, so there is a lot of file activity on this partition. Recently
    I have started getting situations where for no apparent reason, the most
    recently used files are getting corrupted by the later part of the file
    being replaced by garbage. I can manually repair the damaged files and
    carry on from there successfully... for a while, and then the prblem
    occurs again. I think that's the essentials of the scenario.
    Anyway, eventually enough was enough, and as the partition was several
    years old, I reformatted and restored the partition. Everything has been
    alright so far.
    It's a bit worrying - suppose a file got damaged and was undetected
    right through the back-up cycles.
    I can't find any reference to this as a known problem and so I don't
    know how to best deal with it. Has anyone any comments?
    Regards
    Peter

  2. Re: FIle corruption on HPFS partition

    Peter J Seymour wrote:
    > On a FP15 system, I have an HPFS partition containing numerous small
    > files (java source files). These files are frequently edited and
    > compiled, so there is a lot of file activity on this partition. Recently
    > I have started getting situations where for no apparent reason, the most
    > recently used files are getting corrupted by the later part of the file
    > being replaced by garbage. I can manually repair the damaged files and
    > carry on from there successfully... for a while, and then the prblem
    > occurs again. I think that's the essentials of the scenario.


    No system crashes before?

    > Anyway, eventually enough was enough, and as the partition was several
    > years old, I reformatted and restored the partition. Everything has been
    > alright so far.
    > It's a bit worrying - suppose a file got damaged and was undetected
    > right through the back-up cycles.
    > I can't find any reference to this as a known problem and so I don't
    > know how to best deal with it. Has anyone any comments?


    From my oppinion it looks like a hardware problem. Or at least
    something at really low level, like a bad ring 0 driver.
    HPFS is well known to be rock solid. Of course, it cannot compensate for
    that many kind of chrashes as journaling filesystems can do. But it is
    still very reliable. I use it since 1992 and I reformatted the system
    partition only when I changed the HDD. About 4 times so far. And I have
    many source files on the disk too. Ocasionally I had some million files
    on the volume (freedb expiriments). Now there are only 160,000.

    Most of the very supicious problems with OS/2 in my surrounding were
    sooner or later identified to be hardware problems. And there were very
    supicious ones. Everything works, except for the compilation of one
    source file that failed (100% reproducable). Or hangs at very different
    steps of boot, but once the system is up everything is fine.
    Once the reason was a counterfeit cpu, once the electrolytics on the
    mainboard were blown, once the SCSI cable was demaged, once a SS7 board
    did not operate reliable when running at 100MHz FSB, once the power
    supply unit got a bit unstable and so on.

    Unfortunately this won't help you much. Maybe once you will get the
    reason. I can't recommend more than to use a repository for your source
    files located on a network share frequently.


    Marcel

  3. Re: FIle corruption on HPFS partition

    In <4659f16f$0$6441$9b4e6d93@newsspool2.arcor-online.net>, on 05/27/2007
    at 11:00 PM, Marcel Müller said:



    >Peter J Seymour wrote:
    >> On a FP15 system, I have an HPFS partition containing numerous small
    >> files (java source files). These files are frequently edited and
    >> compiled, so there is a lot of file activity on this partition. Recently
    >> I have started getting situations where for no apparent reason, the most
    >> recently used files are getting corrupted by the later part of the file
    >> being replaced by garbage. I can manually repair the damaged files and
    >> carry on from there successfully... for a while, and then the prblem
    >> occurs again. I think that's the essentials of the scenario.


    >No system crashes before?



    That is my question too. HPFS does not corrupt files. Something else is
    taking place too.


    >> Anyway, eventually enough was enough, and as the partition was several
    >> years old, I reformatted and restored the partition. Everything has been
    >> alright so far.
    >> It's a bit worrying - suppose a file got damaged and was undetected
    >> right through the back-up cycles.
    >> I can't find any reference to this as a known problem and so I don't
    >> know how to best deal with it. Has anyone any comments?


    > From my oppinion it looks like a hardware problem. Or at least
    >something at really low level, like a bad ring 0 driver. HPFS is well
    >known to be rock solid. Of course, it cannot compensate for that many
    >kind of chrashes as journaling filesystems can do. But it is still very
    >reliable. I use it since 1992 and I reformatted the system partition
    >only when I changed the HDD. About 4 times so far. And I have many
    >source files on the disk too. Ocasionally I had some million files on
    >the volume (freedb expiriments). Now there are only 160,000.


    >Most of the very supicious problems with OS/2 in my surrounding were
    >sooner or later identified to be hardware problems. And there were very
    >supicious ones. Everything works, except for the compilation of one
    >source file that failed (100% reproducable). Or hangs at very different
    >steps of boot, but once the system is up everything is fine. Once the
    >reason was a counterfeit cpu, once the electrolytics on the mainboard
    >were blown, once the SCSI cable was demaged, once a SS7 board did not
    >operate reliable when running at 100MHz FSB, once the power supply unit
    >got a bit unstable and so on.


    >Unfortunately this won't help you much. Maybe once you will get the
    >reason. I can't recommend more than to use a repository for your source
    >files located on a network share frequently.



    >Marcel



  4. Re: FIle corruption on HPFS partition

    letoured@nospam.net wrote:
    > In <4659f16f$0$6441$9b4e6d93@newsspool2.arcor-online.net>, on 05/27/2007
    > at 11:00 PM, Marcel Müller said:
    >
    >
    >
    >
    >>Peter J Seymour wrote:
    >>
    >>>On a FP15 system, I have an HPFS partition containing numerous small
    >>>files (java source files). These files are frequently edited and
    >>>compiled, so there is a lot of file activity on this partition. Recently
    >>>I have started getting situations where for no apparent reason, the most
    >>>recently used files are getting corrupted by the later part of the file
    >>>being replaced by garbage. I can manually repair the damaged files and
    >>>carry on from there successfully... for a while, and then the prblem
    >>>occurs again. I think that's the essentials of the scenario.

    >
    >
    >>No system crashes before?

    >
    >
    >
    > That is my question too. HPFS does not corrupt files. Something else is
    > taking place too.
    >
    >
    >
    >>>Anyway, eventually enough was enough, and as the partition was several
    >>>years old, I reformatted and restored the partition. Everything has been
    >>>alright so far.
    >>>It's a bit worrying - suppose a file got damaged and was undetected
    >>>right through the back-up cycles.
    >>>I can't find any reference to this as a known problem and so I don't
    >>>know how to best deal with it. Has anyone any comments?

    >
    >
    >>From my oppinion it looks like a hardware problem. Or at least
    >>something at really low level, like a bad ring 0 driver. HPFS is well
    >>known to be rock solid. Of course, it cannot compensate for that many
    >>kind of chrashes as journaling filesystems can do. But it is still very
    >>reliable. I use it since 1992 and I reformatted the system partition
    >>only when I changed the HDD. About 4 times so far. And I have many
    >>source files on the disk too. Ocasionally I had some million files on
    >>the volume (freedb expiriments). Now there are only 160,000.

    >
    >
    >>Most of the very supicious problems with OS/2 in my surrounding were
    >>sooner or later identified to be hardware problems. And there were very
    >>supicious ones. Everything works, except for the compilation of one
    >>source file that failed (100% reproducable). Or hangs at very different
    >>steps of boot, but once the system is up everything is fine. Once the
    >>reason was a counterfeit cpu, once the electrolytics on the mainboard
    >>were blown, once the SCSI cable was demaged, once a SS7 board did not
    >>operate reliable when running at 100MHz FSB, once the power supply unit
    >>got a bit unstable and so on.

    >
    >
    >>Unfortunately this won't help you much. Maybe once you will get the
    >>reason. I can't recommend more than to use a repository for your source
    >>files located on a network share frequently.

    >
    >
    >
    >>Marcel

    >
    >

    I'm still getting the problem after the partition re-format, so I'm
    inclined to agree that the HPFS file system is not at fault.
    On the hardware front - it's a thinkpad with a dead battery, so it's
    running only on mains power. Could that really be the cause of the
    problem? As to the hard disk, it's a travelstar abount four years old.
    As far as I am aware, the problem affects writing to disk but not
    reading. The machine in question is already relegated to being a
    secondary machine, so it might just get completely retired sooner than
    expected.
    Regards
    Peter

  5. Re: FIle corruption on HPFS partition

    Peter J Seymour wrote:
    > On the hardware front - it's a thinkpad with a dead battery, so it's
    > running only on mains power. Could that really be the cause of the
    > problem?


    Well, if the battery is really dead, there may be problems whith the
    power supply. Not that the machine needs the battery, but I heard from
    someone who had problems when moving the notebook. It turned out, that
    probably the supply connection was not sufficiently stable when the plug
    is rotated. Something like that I mean.

    > As to the hard disk, it's a travelstar abount four years old.
    > As far as I am aware, the problem affects writing to disk but not
    > reading.


    It might be some kind of reset to the drive. This may be introduced
    either by power supply problems (including the HDD connector) or even by
    the host, if there are communication failures. Some hard disks forget
    their write back cache when a bus reset arrives.
    You could try to disable the write back cache of the HDD to proove for this.


    > The machine in question is already relegated to being a
    > secondary machine, so it might just get completely retired sooner than
    > expected.


    Unfortunately this does usually only replace the known problems by other
    ones. Especially when talking about OS/2 on new portable hardware.


    Marcel

  6. Re: FIle corruption on HPFS partition

    On Mon, 28 May 2007 20:21:22 UTC in comp.os.os2.misc, Peter J Seymour
    wrote:

    > 'm still getting the problem after the partition re-format, so I'm
    > inclined to agree that the HPFS file system is not at fault.
    > On the hardware front - it's a thinkpad with a dead battery, so it's
    > running only on mains power. Could that really be the cause of the
    > problem? As to the hard disk, it's a travelstar abount four years old.


    I think I've seen something similar before on a soon-to-be-ex-hard disk. I
    doubti f there's a way to prove this without watching it fail though.

    --
    Trevor Hemsley, Brighton, UK
    Trevor dot Hemsley at ntlworld dot com

  7. Re: FIle corruption on HPFS partition

    Trevor Hemsley wrote:
    > On Mon, 28 May 2007 20:21:22 UTC in comp.os.os2.misc, Peter J Seymour
    > wrote:
    >
    >
    >>'m still getting the problem after the partition re-format, so I'm
    >>inclined to agree that the HPFS file system is not at fault.
    >>On the hardware front - it's a thinkpad with a dead battery, so it's
    >>running only on mains power. Could that really be the cause of the
    >>problem? As to the hard disk, it's a travelstar abount four years old.

    >
    >
    > I think I've seen something similar before on a soon-to-be-ex-hard disk. I
    > doubti f there's a way to prove this without watching it fail though.
    >

    I have a spare unused hard disk, so I will try that (prefereable to
    abandoning a whole thinkpad). Fortunately I have a holder for fitting a
    second disk, so I can just xcopy the os/2 partitions. I will worry about
    re-instating windows later.
    I expect to report back in a few days time.
    Regards
    Peter

  8. Re: FIle corruption on HPFS partition

    Marcel Müller wrote:
    > Peter J Seymour wrote:
    >
    >> On the hardware front - it's a thinkpad with a dead battery, so it's
    >> running only on mains power. Could that really be the cause of the
    >> problem?

    >
    >
    > Well, if the battery is really dead, there may be problems whith the
    > power supply. Not that the machine needs the battery, but I heard from
    > someone who had problems when moving the notebook. It turned out, that
    > probably the supply connection was not sufficiently stable when the plug
    > is rotated. Something like that I mean.
    >
    >> As to the hard disk, it's a travelstar abount four years old.
    >> As far as I am aware, the problem affects writing to disk but not
    >> reading.

    >
    >
    > It might be some kind of reset to the drive. This may be introduced
    > either by power supply problems (including the HDD connector) or even by
    > the host, if there are communication failures. Some hard disks forget
    > their write back cache when a bus reset arrives.
    > You could try to disable the write back cache of the HDD to proove for
    > this.
    >
    >
    >> The machine in question is already relegated to being a secondary
    >> machine, so it might just get completely retired sooner than expected.

    >
    >
    > Unfortunately this does usually only replace the known problems by other
    > ones. Especially when talking about OS/2 on new portable hardware.
    >
    >
    > Marcel

    If all else fails, it may come down to the choice of forking out for a
    new battery or ditching the whole machine.
    Regards
    Peter

  9. Re: FIle corruption on HPFS partition

    Peter J Seymour schrieb:
    >
    > letoured@nospam.net wrote:
    > > In <4659f16f$0$6441$9b4e6d93@newsspool2.arcor-online.net>, on 05/27/2007
    > > at 11:00 PM, Marcel Müller said:
    > >
    > >
    > >
    > >
    > >>Peter J Seymour wrote:
    > >>
    > >>>On a FP15 system, I have an HPFS partition containing numerous small
    > >>>files (java source files). These files are frequently edited and
    > >>>compiled, so there is a lot of file activity on this partition. Recently
    > >>>I have started getting situations where for no apparent reason, the most
    > >>>recently used files are getting corrupted by the later part of the file
    > >>>being replaced by garbage. I can manually repair the damaged files and
    > >>>carry on from there successfully... for a while, and then the prblem
    > >>>occurs again. I think that's the essentials of the scenario.

    > >
    > >
    > >>No system crashes before?

    > >
    > >
    > >
    > > That is my question too. HPFS does not corrupt files. Something else is
    > > taking place too.
    > >
    > >
    > >
    > >>>Anyway, eventually enough was enough, and as the partition was several
    > >>>years old, I reformatted and restored the partition. Everything has been
    > >>>alright so far.
    > >>>It's a bit worrying - suppose a file got damaged and was undetected
    > >>>right through the back-up cycles.
    > >>>I can't find any reference to this as a known problem and so I don't
    > >>>know how to best deal with it. Has anyone any comments?

    > >
    > >
    > >>From my oppinion it looks like a hardware problem. Or at least
    > >>something at really low level, like a bad ring 0 driver. HPFS is well
    > >>known to be rock solid. Of course, it cannot compensate for that many
    > >>kind of chrashes as journaling filesystems can do. But it is still very
    > >>reliable. I use it since 1992 and I reformatted the system partition
    > >>only when I changed the HDD. About 4 times so far. And I have many
    > >>source files on the disk too. Ocasionally I had some million files on
    > >>the volume (freedb expiriments). Now there are only 160,000.

    > >
    > >
    > >>Most of the very supicious problems with OS/2 in my surrounding were
    > >>sooner or later identified to be hardware problems. And there were very
    > >>supicious ones. Everything works, except for the compilation of one
    > >>source file that failed (100% reproducable). Or hangs at very different
    > >>steps of boot, but once the system is up everything is fine. Once the
    > >>reason was a counterfeit cpu, once the electrolytics on the mainboard
    > >>were blown, once the SCSI cable was demaged, once a SS7 board did not
    > >>operate reliable when running at 100MHz FSB, once the power supply unit
    > >>got a bit unstable and so on.

    > >
    > >
    > >>Unfortunately this won't help you much. Maybe once you will get the
    > >>reason. I can't recommend more than to use a repository for your source
    > >>files located on a network share frequently.

    > >
    > >
    > >
    > >>Marcel

    > >
    > >

    > I'm still getting the problem after the partition re-format, so I'm
    > inclined to agree that the HPFS file system is not at fault.
    > On the hardware front - it's a thinkpad with a dead battery, so it's
    > running only on mains power. Could that really be the cause of the
    > problem? As to the hard disk, it's a travelstar abount four years old.
    > As far as I am aware, the problem affects writing to disk but not
    > reading. The machine in question is already relegated to being a
    > secondary machine, so it might just get completely retired sooner than
    > expected.


    Some manucatours of hard disks offer programs to test their disk.
    Maybe you should execute the according one and see what it says.

    Hendrik

  10. Re: FIle corruption on HPFS partition

    Peter J Seymour wrote:
    > On a FP15 system, I have an HPFS partition containing numerous small
    > files (java source files). These files are frequently edited and
    > compiled, so there is a lot of file activity on this partition. Recently
    > I have started getting situations where for no apparent reason, the most
    > recently used files are getting corrupted by the later part of the file
    > being replaced by garbage. I can manually repair the damaged files and
    > carry on from there successfully... for a while, and then the prblem
    > occurs again. I think that's the essentials of the scenario.
    > Anyway, eventually enough was enough, and as the partition was several
    > years old, I reformatted and restored the partition. Everything has been
    > alright so far.
    > It's a bit worrying - suppose a file got damaged and was undetected
    > right through the back-up cycles.
    > I can't find any reference to this as a known problem and so I don't
    > know how to best deal with it. Has anyone any comments?
    > Regards
    > Peter

    Finally returning to this topic. First of all, my thanks for the various
    suggestions made. Nothing has directly solved the problem, but that's
    life I suppose.
    Two useful things emerged:
    - I knew that disk analysis programs existed but had the impression that
    they were difficult to come by. However, Hitachi have a 'Disk Fitness'
    utility available for download which seems to have come out in 2004, so
    it goes to show one can miss these things. It gave the disk in question
    a clean bill of health.
    - Thinkpad batteries seem to be very much cheaper that when I last looked
    so I felt less aggrieved in buying one.
    Anyway, after experimentation without fixing the problem, I noticed that
    the problem mostly seemed to occur late afternoon/early evening. This is
    when household appliances are most likely to be running, so this
    suggests interference on the mains supply as a possible culprit.
    I have decided to try a policy of running the thinkpad on mains
    electricity in the main part of the day and on battery (using the newly
    purchased working one) later on. Another sensible move would be to fit a
    high value suppressor on the mains.
    I'm not yet totally convinved that this fixes/worksaround the problem,
    but it looks promising.

    Peter

+ Reply to Thread