system page file help..

Discussion in 'Computer Information' started by Robert Baer, Jan 5, 2013.

  1. Robert Baer

    Robert Baer Guest

    This very rarely happens to me, but when it does, it is a total (to
    date) non-recoverable bitch.

    Background: my HD has the following structure:
    Partition 1 = C: WIN9SE (primary)
    Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    Partition 3 = F: MASTER (primary)
    Partition 4 = G: WIN2K (primary).
    My backup drive is exactly the same.

    Somehow i did something when both drives were plugged into the
    computer (i use removable HD kits for all hard drives), and now when i
    boot with only one HD, i get this incredibly stupid and insane message
    box "your system has no paging file, or the paging file is too small..."
    and it goes on, ASS-U-MEing one is fully booted, with the procedure of
    setting the paging file.

    It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    anything other that unplug the computer to turn it off.
    There is absolutely NO F*ING WAY to change the paging file(s) unless
    one _then_ boots with both hard drives in the computer.

    And wasting hours with various page file allocations makes no
    difference in the above "operation".

    Please, ANY clues as what else one can do to try and fix this?

    PS:
    Do not mention useless "safe mode"; one gets a black screen with
    "safe mode" in all 4 corners and the cursor, and nothing else (oh yes,
    one must unplug computer to reset).
    This is an artifact when one has a Windoze dual-boot system; i
    experimented one time without Win98SE (Win2K only OS, boots directly)
    and sure enough, "safe mode" is actually useable; i damn near fainted
    due to the shock (many years of useless "safe mode" with a dual-boot
    system).
    Robert Baer, Jan 5, 2013
    #1
    1. Advertising

  2. Robert Baer

    Paul Guest

    Robert Baer wrote:
    > This very rarely happens to me, but when it does, it is a total (to
    > date) non-recoverable bitch.
    >
    > Background: my HD has the following structure:
    > Partition 1 = C: WIN9SE (primary)
    > Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    > Partition 3 = F: MASTER (primary)
    > Partition 4 = G: WIN2K (primary).
    > My backup drive is exactly the same.
    >
    > Somehow i did something when both drives were plugged into the
    > computer (i use removable HD kits for all hard drives), and now when i
    > boot with only one HD, i get this incredibly stupid and insane message
    > box "your system has no paging file, or the paging file is too small..."
    > and it goes on, ASS-U-MEing one is fully booted, with the procedure of
    > setting the paging file.
    >
    > It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    > anything other that unplug the computer to turn it off.
    > There is absolutely NO F*ING WAY to change the paging file(s) unless
    > one _then_ boots with both hard drives in the computer.
    >
    > And wasting hours with various page file allocations makes no
    > difference in the above "operation".
    >
    > Please, ANY clues as what else one can do to try and fix this?
    >
    > PS:
    > Do not mention useless "safe mode"; one gets a black screen with "safe
    > mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    > must unplug computer to reset).
    > This is an artifact when one has a Windoze dual-boot system; i
    > experimented one time without Win98SE (Win2K only OS, boots directly)
    > and sure enough, "safe mode" is actually useable; i damn near fainted
    > due to the shock (many years of useless "safe mode" with a dual-boot
    > system).


    I had something like that happen with Win2K.

    And the funny part was, even putting a pagefile file onto C:,
    didn't fix it. I even copied a pagefile from another OS,
    thinking that perhaps the pagefile had a header and some info
    that needed to be present, to be recognized.

    Which implies some registry info needs to be doctored as well.
    Maybe something like that would have fixed it. (You may be
    able to edit the registry, from another OS.)

    I suppose a Repair Install could fix it. As I was in a hurry at the
    time (and was working in a VM), I just started over again, rather
    than waste the time slugging at it.

    You could also look at what registry repair options you have
    available. System Restore keeps copies of the registry,
    but I don't think there is System Restore on the older OSes.

    Post back if you find a solution.

    Paul
    Paul, Jan 5, 2013
    #2
    1. Advertising

  3. Robert Baer

    Robert Baer Guest

    Paul wrote:
    > Robert Baer wrote:
    >> This very rarely happens to me, but when it does, it is a total (to
    >> date) non-recoverable bitch.
    >>
    >> Background: my HD has the following structure:
    >> Partition 1 = C: WIN9SE (primary)
    >> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >> Partition 3 = F: MASTER (primary)
    >> Partition 4 = G: WIN2K (primary).
    >> My backup drive is exactly the same.
    >>
    >> Somehow i did something when both drives were plugged into the
    >> computer (i use removable HD kits for all hard drives), and now when i
    >> boot with only one HD, i get this incredibly stupid and insane message
    >> box "your system has no paging file, or the paging file is too
    >> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >> procedure of setting the paging file.
    >>
    >> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >> anything other that unplug the computer to turn it off.
    >> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >> one _then_ boots with both hard drives in the computer.
    >>
    >> And wasting hours with various page file allocations makes no
    >> difference in the above "operation".
    >>
    >> Please, ANY clues as what else one can do to try and fix this?
    >>
    >> PS:
    >> Do not mention useless "safe mode"; one gets a black screen with "safe
    >> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >> must unplug computer to reset).
    >> This is an artifact when one has a Windoze dual-boot system; i
    >> experimented one time without Win98SE (Win2K only OS, boots directly)
    >> and sure enough, "safe mode" is actually useable; i damn near fainted
    >> due to the shock (many years of useless "safe mode" with a dual-boot
    >> system).

    >
    > I had something like that happen with Win2K.
    >
    > And the funny part was, even putting a pagefile file onto C:,
    > didn't fix it. I even copied a pagefile from another OS,
    > thinking that perhaps the pagefile had a header and some info
    > that needed to be present, to be recognized.
    >
    > Which implies some registry info needs to be doctored as well.
    > Maybe something like that would have fixed it. (You may be
    > able to edit the registry, from another OS.)
    >
    > I suppose a Repair Install could fix it. As I was in a hurry at the
    > time (and was working in a VM), I just started over again, rather
    > than waste the time slugging at it.
    >
    > You could also look at what registry repair options you have
    > available. System Restore keeps copies of the registry,
    > but I don't think there is System Restore on the older OSes.
    >
    > Post back if you find a solution.
    >
    > Paul

    "Repair Install" sounds nice, but rarely helps any problem; in this
    case it does nothing.
    Naturally, there are NO registry repair options that i know of.
    And not even Hisself Billy Boy knows where they hide the registry.
    RegEdit only works on the registry for that particular OS on its volume.
    The only way to fix it is to start from scratch like you mentioned.
    And that takes daze to re-build all of the apps.
    Robert Baer, Jan 6, 2013
    #3
  4. Robert Baer

    Paul Guest

    Robert Baer wrote:
    > Paul wrote:
    >> Robert Baer wrote:
    >>> This very rarely happens to me, but when it does, it is a total (to
    >>> date) non-recoverable bitch.
    >>>
    >>> Background: my HD has the following structure:
    >>> Partition 1 = C: WIN9SE (primary)
    >>> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >>> Partition 3 = F: MASTER (primary)
    >>> Partition 4 = G: WIN2K (primary).
    >>> My backup drive is exactly the same.
    >>>
    >>> Somehow i did something when both drives were plugged into the
    >>> computer (i use removable HD kits for all hard drives), and now when i
    >>> boot with only one HD, i get this incredibly stupid and insane message
    >>> box "your system has no paging file, or the paging file is too
    >>> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >>> procedure of setting the paging file.
    >>>
    >>> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >>> anything other that unplug the computer to turn it off.
    >>> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >>> one _then_ boots with both hard drives in the computer.
    >>>
    >>> And wasting hours with various page file allocations makes no
    >>> difference in the above "operation".
    >>>
    >>> Please, ANY clues as what else one can do to try and fix this?
    >>>
    >>> PS:
    >>> Do not mention useless "safe mode"; one gets a black screen with "safe
    >>> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >>> must unplug computer to reset).
    >>> This is an artifact when one has a Windoze dual-boot system; i
    >>> experimented one time without Win98SE (Win2K only OS, boots directly)
    >>> and sure enough, "safe mode" is actually useable; i damn near fainted
    >>> due to the shock (many years of useless "safe mode" with a dual-boot
    >>> system).

    >>
    >> I had something like that happen with Win2K.
    >>
    >> And the funny part was, even putting a pagefile file onto C:,
    >> didn't fix it. I even copied a pagefile from another OS,
    >> thinking that perhaps the pagefile had a header and some info
    >> that needed to be present, to be recognized.
    >>
    >> Which implies some registry info needs to be doctored as well.
    >> Maybe something like that would have fixed it. (You may be
    >> able to edit the registry, from another OS.)
    >>
    >> I suppose a Repair Install could fix it. As I was in a hurry at the
    >> time (and was working in a VM), I just started over again, rather
    >> than waste the time slugging at it.
    >>
    >> You could also look at what registry repair options you have
    >> available. System Restore keeps copies of the registry,
    >> but I don't think there is System Restore on the older OSes.
    >>
    >> Post back if you find a solution.
    >>
    >> Paul

    > "Repair Install" sounds nice, but rarely helps any problem; in this
    > case it does nothing.
    > Naturally, there are NO registry repair options that i know of.
    > And not even Hisself Billy Boy knows where they hide the registry.
    > RegEdit only works on the registry for that particular OS on its volume.
    > The only way to fix it is to start from scratch like you mentioned.
    > And that takes daze to re-build all of the apps.
    >


    This is one I can find in WinXP. It occurs several places.

    HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Session Manager\Memory Management
    PagingFiles c:\pagefile.sys 2048 4092 <--- MultiStrings Type

    HKEY_LOCAL_MACHINE\SYSTEM\ControlSet002\Control\Session Manager\Memory Management
    PagingFiles c:\pagefile.sys 2048 4092 <--- MultiStrings Type

    HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management
    PagingFiles c:\pagefile.sys 2048 4092 <--- MultiStrings Type

    ControlSet002 = "Last Known Good"
    ControlSet001 = last control set booted with
    CurrentControlSet = pointer to one of the other two

    Maybe when I'm more awake, I'll try some experiments in a Win2K VM.

    Paul
    Paul, Jan 6, 2013
    #4
  5. Robert Baer

    Paul Guest

    Robert Baer wrote:
    > Paul wrote:
    >> Robert Baer wrote:
    >>> This very rarely happens to me, but when it does, it is a total (to
    >>> date) non-recoverable bitch.
    >>>
    >>> Background: my HD has the following structure:
    >>> Partition 1 = C: WIN9SE (primary)
    >>> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >>> Partition 3 = F: MASTER (primary)
    >>> Partition 4 = G: WIN2K (primary).
    >>> My backup drive is exactly the same.
    >>>
    >>> Somehow i did something when both drives were plugged into the
    >>> computer (i use removable HD kits for all hard drives), and now when i
    >>> boot with only one HD, i get this incredibly stupid and insane message
    >>> box "your system has no paging file, or the paging file is too
    >>> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >>> procedure of setting the paging file.
    >>>
    >>> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >>> anything other that unplug the computer to turn it off.
    >>> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >>> one _then_ boots with both hard drives in the computer.
    >>>
    >>> And wasting hours with various page file allocations makes no
    >>> difference in the above "operation".
    >>>
    >>> Please, ANY clues as what else one can do to try and fix this?
    >>>
    >>> PS:
    >>> Do not mention useless "safe mode"; one gets a black screen with "safe
    >>> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >>> must unplug computer to reset).
    >>> This is an artifact when one has a Windoze dual-boot system; i
    >>> experimented one time without Win98SE (Win2K only OS, boots directly)
    >>> and sure enough, "safe mode" is actually useable; i damn near fainted
    >>> due to the shock (many years of useless "safe mode" with a dual-boot
    >>> system).

    >>
    >> I had something like that happen with Win2K.
    >>
    >> And the funny part was, even putting a pagefile file onto C:,
    >> didn't fix it. I even copied a pagefile from another OS,
    >> thinking that perhaps the pagefile had a header and some info
    >> that needed to be present, to be recognized.
    >>
    >> Which implies some registry info needs to be doctored as well.
    >> Maybe something like that would have fixed it. (You may be
    >> able to edit the registry, from another OS.)
    >>
    >> I suppose a Repair Install could fix it. As I was in a hurry at the
    >> time (and was working in a VM), I just started over again, rather
    >> than waste the time slugging at it.
    >>
    >> You could also look at what registry repair options you have
    >> available. System Restore keeps copies of the registry,
    >> but I don't think there is System Restore on the older OSes.
    >>
    >> Post back if you find a solution.
    >>
    >> Paul

    > "Repair Install" sounds nice, but rarely helps any problem; in this
    > case it does nothing.
    > Naturally, there are NO registry repair options that i know of.
    > And not even Hisself Billy Boy knows where they hide the registry.
    > RegEdit only works on the registry for that particular OS on its volume.
    > The only way to fix it is to start from scratch like you mentioned.
    > And that takes daze to re-build all of the apps.
    >


    OK. I fired up my virtual machine and did some testing.

    1) I tried a config with just one disk drive, with Win2K C: on it.
    I deleted C:\pagefile.sys. I restarted. The system created a
    new pagefile.sys and carried on as if nothing happened. OK,
    so that's not the recipe on how to break it.

    2) Next, I tried two disks with Win2K on them, a C: on each disk.
    I assigned a label to each disk, so I could tell them apart.
    I also used Sysinternals VolumeID and gave the second disk a
    different volume ID. Just in case that was important. (It's possible
    the OS will do that anyway, but I didn't want to ruin things by
    ignoring it.)

    Now, what happens at boot, is interesting. I expected this to
    happen, with the two disks present.

    Disk0 C:
    Disk1 E:

    And instead, they had switched roles.

    Disk0 E: Healthy (System)
    Disk1 C: Healthy (Pagefile)

    So what's happened is, the system should have booted from Disk0.
    And it did. But, when it did, two things happened. It decided the
    other disk had C:\pagefile.sys (which it does). And then it changes the
    system drive letter to E: ! At this point, no pagefile.sys was
    deleted.

    If I delete pagefile.sys from E:, nothing happens, because it's
    still pointed at Disk1 C:. I can still boot, because both disks
    are still present, and the pagefile.sys on the so-called C: is
    still available.

    OK, now, remove Disk1 and try to run with Disk0. BOom. Now,
    we're in that pagefile error message loop.

    And basically, nothing I did would fix it. I tried changing
    the registry, by editing SYSTEM with a hex editor, and changing
    that registry entry. (Hmmm. It occurs to me, there are three
    similar keys, and maybe I edited the wrong one. Mumble...)

    My conclusion at the moment is, if you were to put back the
    disks that were present when the "ole switcheroo" was pulled,
    then the Win2K system would boot. It would look like my mess,
    with C: having become E: and E: changing to C:. You should put the
    disks back on the same cables they used originally.

    So there are actually two variables involved. There is the
    registry entry that keeps track of system drive letter.
    And also the registry entry that keeps the pagefile strings.
    In fact, the pagefile registry entry isn't changing. It's the
    fact the drive letters have changed, that screws things up.
    And between the two of those, you have to get things aligned
    so the system can find the pagefile.

    Maybe this situation arises, when a person clones Win2K, and
    both disks with their copy of Win2K remain present, for the very
    first boot of the clone. Then the drive letter mess happens
    and so on. And then, disconnecting the second disk, and trying
    to boot with only one of the two disks present, makes it
    painfully obvious that the pagefile is missing as far
    as the system is concerned.

    Paul
    Paul, Jan 7, 2013
    #5
  6. Robert Baer

    Robert Baer Guest

    Paul wrote:
    > Robert Baer wrote:
    >> This very rarely happens to me, but when it does, it is a total (to
    >> date) non-recoverable bitch.
    >>
    >> Background: my HD has the following structure:
    >> Partition 1 = C: WIN9SE (primary)
    >> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >> Partition 3 = F: MASTER (primary)
    >> Partition 4 = G: WIN2K (primary).
    >> My backup drive is exactly the same.
    >>
    >> Somehow i did something when both drives were plugged into the
    >> computer (i use removable HD kits for all hard drives), and now when i
    >> boot with only one HD, i get this incredibly stupid and insane message
    >> box "your system has no paging file, or the paging file is too
    >> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >> procedure of setting the paging file.
    >>
    >> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >> anything other that unplug the computer to turn it off.
    >> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >> one _then_ boots with both hard drives in the computer.
    >>
    >> And wasting hours with various page file allocations makes no
    >> difference in the above "operation".
    >>
    >> Please, ANY clues as what else one can do to try and fix this?
    >>
    >> PS:
    >> Do not mention useless "safe mode"; one gets a black screen with "safe
    >> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >> must unplug computer to reset).
    >> This is an artifact when one has a Windoze dual-boot system; i
    >> experimented one time without Win98SE (Win2K only OS, boots directly)
    >> and sure enough, "safe mode" is actually useable; i damn near fainted
    >> due to the shock (many years of useless "safe mode" with a dual-boot
    >> system).

    >
    > I had something like that happen with Win2K.
    >
    > And the funny part was, even putting a pagefile file onto C:,
    > didn't fix it. I even copied a pagefile from another OS,
    > thinking that perhaps the pagefile had a header and some info
    > that needed to be present, to be recognized.
    >
    > Which implies some registry info needs to be doctored as well.
    > Maybe something like that would have fixed it. (You may be
    > able to edit the registry, from another OS.)
    >
    > I suppose a Repair Install could fix it. As I was in a hurry at the
    > time (and was working in a VM), I just started over again, rather
    > than waste the time slugging at it.
    >
    > You could also look at what registry repair options you have
    > available. System Restore keeps copies of the registry,
    > but I don't think there is System Restore on the older OSes.
    >
    > Post back if you find a solution.
    >
    > Paul

    At present, i cannot test this possible way to fix that problem
    (would need a spare HD to make new exclusive install of Win2K): at boot,
    F8 to "Safe Mode" and try the control panel / system / advanced /
    performance options / virtual memory / change..
    Robert Baer, Jan 8, 2013
    #6
  7. Robert Baer

    Robert Baer Guest

    Paul wrote:
    > Robert Baer wrote:
    >> Paul wrote:
    >>> Robert Baer wrote:
    >>>> This very rarely happens to me, but when it does, it is a total (to
    >>>> date) non-recoverable bitch.
    >>>>
    >>>> Background: my HD has the following structure:
    >>>> Partition 1 = C: WIN9SE (primary)
    >>>> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >>>> Partition 3 = F: MASTER (primary)
    >>>> Partition 4 = G: WIN2K (primary).
    >>>> My backup drive is exactly the same.
    >>>>
    >>>> Somehow i did something when both drives were plugged into the
    >>>> computer (i use removable HD kits for all hard drives), and now when i
    >>>> boot with only one HD, i get this incredibly stupid and insane message
    >>>> box "your system has no paging file, or the paging file is too
    >>>> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >>>> procedure of setting the paging file.
    >>>>
    >>>> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >>>> anything other that unplug the computer to turn it off.
    >>>> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >>>> one _then_ boots with both hard drives in the computer.
    >>>>
    >>>> And wasting hours with various page file allocations makes no
    >>>> difference in the above "operation".
    >>>>
    >>>> Please, ANY clues as what else one can do to try and fix this?
    >>>>
    >>>> PS:
    >>>> Do not mention useless "safe mode"; one gets a black screen with "safe
    >>>> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >>>> must unplug computer to reset).
    >>>> This is an artifact when one has a Windoze dual-boot system; i
    >>>> experimented one time without Win98SE (Win2K only OS, boots directly)
    >>>> and sure enough, "safe mode" is actually useable; i damn near fainted
    >>>> due to the shock (many years of useless "safe mode" with a dual-boot
    >>>> system).
    >>>
    >>> I had something like that happen with Win2K.
    >>>
    >>> And the funny part was, even putting a pagefile file onto C:,
    >>> didn't fix it. I even copied a pagefile from another OS,
    >>> thinking that perhaps the pagefile had a header and some info
    >>> that needed to be present, to be recognized.
    >>>
    >>> Which implies some registry info needs to be doctored as well.
    >>> Maybe something like that would have fixed it. (You may be
    >>> able to edit the registry, from another OS.)
    >>>
    >>> I suppose a Repair Install could fix it. As I was in a hurry at the
    >>> time (and was working in a VM), I just started over again, rather
    >>> than waste the time slugging at it.
    >>>
    >>> You could also look at what registry repair options you have
    >>> available. System Restore keeps copies of the registry,
    >>> but I don't think there is System Restore on the older OSes.
    >>>
    >>> Post back if you find a solution.
    >>>
    >>> Paul

    >> "Repair Install" sounds nice, but rarely helps any problem; in this
    >> case it does nothing.
    >> Naturally, there are NO registry repair options that i know of.
    >> And not even Hisself Billy Boy knows where they hide the registry.
    >> RegEdit only works on the registry for that particular OS on its volume.
    >> The only way to fix it is to start from scratch like you mentioned.
    >> And that takes daze to re-build all of the apps.
    >>

    >
    > OK. I fired up my virtual machine and did some testing.
    >
    > 1) I tried a config with just one disk drive, with Win2K C: on it.
    > I deleted C:\pagefile.sys. I restarted. The system created a
    > new pagefile.sys and carried on as if nothing happened. OK,
    > so that's not the recipe on how to break it.
    >
    > 2) Next, I tried two disks with Win2K on them, a C: on each disk.
    > I assigned a label to each disk, so I could tell them apart.
    > I also used Sysinternals VolumeID and gave the second disk a
    > different volume ID. Just in case that was important. (It's possible
    > the OS will do that anyway, but I didn't want to ruin things by
    > ignoring it.)
    >
    > Now, what happens at boot, is interesting. I expected this to
    > happen, with the two disks present.
    >
    > Disk0 C:
    > Disk1 E:
    >
    > And instead, they had switched roles.
    >
    > Disk0 E: Healthy (System)
    > Disk1 C: Healthy (Pagefile)
    >
    > So what's happened is, the system should have booted from Disk0.
    > And it did. But, when it did, two things happened. It decided the
    > other disk had C:\pagefile.sys (which it does). And then it changes the
    > system drive letter to E: ! At this point, no pagefile.sys was
    > deleted.
    >
    > If I delete pagefile.sys from E:, nothing happens, because it's
    > still pointed at Disk1 C:. I can still boot, because both disks
    > are still present, and the pagefile.sys on the so-called C: is
    > still available.
    >
    > OK, now, remove Disk1 and try to run with Disk0. BOom. Now,
    > we're in that pagefile error message loop.
    >
    > And basically, nothing I did would fix it. I tried changing
    > the registry, by editing SYSTEM with a hex editor, and changing
    > that registry entry. (Hmmm. It occurs to me, there are three
    > similar keys, and maybe I edited the wrong one. Mumble...)
    >
    > My conclusion at the moment is, if you were to put back the
    > disks that were present when the "ole switcheroo" was pulled,
    > then the Win2K system would boot. It would look like my mess,
    > with C: having become E: and E: changing to C:. You should put the
    > disks back on the same cables they used originally.
    >
    > So there are actually two variables involved. There is the
    > registry entry that keeps track of system drive letter.
    > And also the registry entry that keeps the pagefile strings.
    > In fact, the pagefile registry entry isn't changing. It's the
    > fact the drive letters have changed, that screws things up.
    > And between the two of those, you have to get things aligned
    > so the system can find the pagefile.
    >
    > Maybe this situation arises, when a person clones Win2K, and
    > both disks with their copy of Win2K remain present, for the very
    > first boot of the clone. Then the drive letter mess happens
    > and so on. And then, disconnecting the second disk, and trying
    > to boot with only one of the two disks present, makes it
    > painfully obvious that the pagefile is missing as far
    > as the system is concerned.
    >
    > Paul

    WOW! You found the secret a switcheroo!
    I did not know of that ages ago, what i did the next build was make
    sure Win2K was "out there" (eg: partition 4, G:), and for a number of
    years that prevented the problem (gotta say "somehow" now that i have a
    better idea about the funny business).
    Robert Baer, Jan 8, 2013
    #7
  8. Robert Baer

    Robert Baer Guest

    Paul wrote:
    > Robert Baer wrote:
    >> Paul wrote:
    >>> Robert Baer wrote:
    >>>> This very rarely happens to me, but when it does, it is a total (to
    >>>> date) non-recoverable bitch.
    >>>>
    >>>> Background: my HD has the following structure:
    >>>> Partition 1 = C: WIN9SE (primary)
    >>>> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >>>> Partition 3 = F: MASTER (primary)
    >>>> Partition 4 = G: WIN2K (primary).
    >>>> My backup drive is exactly the same.
    >>>>
    >>>> Somehow i did something when both drives were plugged into the
    >>>> computer (i use removable HD kits for all hard drives), and now when i
    >>>> boot with only one HD, i get this incredibly stupid and insane message
    >>>> box "your system has no paging file, or the paging file is too
    >>>> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >>>> procedure of setting the paging file.
    >>>>
    >>>> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >>>> anything other that unplug the computer to turn it off.
    >>>> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >>>> one _then_ boots with both hard drives in the computer.
    >>>>
    >>>> And wasting hours with various page file allocations makes no
    >>>> difference in the above "operation".
    >>>>
    >>>> Please, ANY clues as what else one can do to try and fix this?
    >>>>
    >>>> PS:
    >>>> Do not mention useless "safe mode"; one gets a black screen with "safe
    >>>> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >>>> must unplug computer to reset).
    >>>> This is an artifact when one has a Windoze dual-boot system; i
    >>>> experimented one time without Win98SE (Win2K only OS, boots directly)
    >>>> and sure enough, "safe mode" is actually useable; i damn near fainted
    >>>> due to the shock (many years of useless "safe mode" with a dual-boot
    >>>> system).
    >>>
    >>> I had something like that happen with Win2K.
    >>>
    >>> And the funny part was, even putting a pagefile file onto C:,
    >>> didn't fix it. I even copied a pagefile from another OS,
    >>> thinking that perhaps the pagefile had a header and some info
    >>> that needed to be present, to be recognized.
    >>>
    >>> Which implies some registry info needs to be doctored as well.
    >>> Maybe something like that would have fixed it. (You may be
    >>> able to edit the registry, from another OS.)
    >>>
    >>> I suppose a Repair Install could fix it. As I was in a hurry at the
    >>> time (and was working in a VM), I just started over again, rather
    >>> than waste the time slugging at it.
    >>>
    >>> You could also look at what registry repair options you have
    >>> available. System Restore keeps copies of the registry,
    >>> but I don't think there is System Restore on the older OSes.
    >>>
    >>> Post back if you find a solution.
    >>>
    >>> Paul

    >> "Repair Install" sounds nice, but rarely helps any problem; in this
    >> case it does nothing.
    >> Naturally, there are NO registry repair options that i know of.
    >> And not even Hisself Billy Boy knows where they hide the registry.
    >> RegEdit only works on the registry for that particular OS on its volume.
    >> The only way to fix it is to start from scratch like you mentioned.
    >> And that takes daze to re-build all of the apps.
    >>

    >
    > OK. I fired up my virtual machine and did some testing.
    >
    > 1) I tried a config with just one disk drive, with Win2K C: on it.
    > I deleted C:\pagefile.sys. I restarted. The system created a
    > new pagefile.sys and carried on as if nothing happened. OK,
    > so that's not the recipe on how to break it.
    >
    > 2) Next, I tried two disks with Win2K on them, a C: on each disk.
    > I assigned a label to each disk, so I could tell them apart.
    > I also used Sysinternals VolumeID and gave the second disk a
    > different volume ID. Just in case that was important. (It's possible
    > the OS will do that anyway, but I didn't want to ruin things by
    > ignoring it.)
    >
    > Now, what happens at boot, is interesting. I expected this to
    > happen, with the two disks present.
    >
    > Disk0 C:
    > Disk1 E:
    >
    > And instead, they had switched roles.
    >
    > Disk0 E: Healthy (System)
    > Disk1 C: Healthy (Pagefile)
    >
    > So what's happened is, the system should have booted from Disk0.
    > And it did. But, when it did, two things happened. It decided the
    > other disk had C:\pagefile.sys (which it does). And then it changes the
    > system drive letter to E: ! At this point, no pagefile.sys was
    > deleted.
    >
    > If I delete pagefile.sys from E:, nothing happens, because it's
    > still pointed at Disk1 C:. I can still boot, because both disks
    > are still present, and the pagefile.sys on the so-called C: is
    > still available.
    >
    > OK, now, remove Disk1 and try to run with Disk0. BOom. Now,
    > we're in that pagefile error message loop.
    >
    > And basically, nothing I did would fix it. I tried changing
    > the registry, by editing SYSTEM with a hex editor, and changing
    > that registry entry. (Hmmm. It occurs to me, there are three
    > similar keys, and maybe I edited the wrong one. Mumble...)
    >
    > My conclusion at the moment is, if you were to put back the
    > disks that were present when the "ole switcheroo" was pulled,
    > then the Win2K system would boot. It would look like my mess,
    > with C: having become E: and E: changing to C:. You should put the
    > disks back on the same cables they used originally.
    >
    > So there are actually two variables involved. There is the
    > registry entry that keeps track of system drive letter.
    > And also the registry entry that keeps the pagefile strings.
    > In fact, the pagefile registry entry isn't changing. It's the
    > fact the drive letters have changed, that screws things up.
    > And between the two of those, you have to get things aligned
    > so the system can find the pagefile.
    >
    > Maybe this situation arises, when a person clones Win2K, and
    > both disks with their copy of Win2K remain present, for the very
    > first boot of the clone. Then the drive letter mess happens
    > and so on. And then, disconnecting the second disk, and trying
    > to boot with only one of the two disks present, makes it
    > painfully obvious that the pagefile is missing as far
    > as the system is concerned.
    >
    > Paul

    I think i have found a way to reliably create the problem as well as
    fix it.
    Use Acronis to clone HD1 to HD2 and re-boot with both connected.

    Original HD1 and what should be if all OK for HD2
    Partition 1 = C: WIN9SE (primary) I:
    Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    Partition 3 = F: MASTER (primary) L:
    Partition 4 = G: WIN2K (primary). M:

    But, you get C: D:
    E: & F: G: & H:
    I: J:
    ?: ?: (did not look)
    Ugly results were reported by DOS Fdisk.
    *
    Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    need to use (say) P: as temp placeholder; final result to look like the
    initial sample (ie: BE IN ORDER on each drive).
    Robert Baer, Jan 10, 2013
    #8
  9. Robert Baer

    Guest

    Robert Baer wrote:

    <Snip>
    >> Paul

    > I think i have found a way to reliably create the problem as well as
    > fix it.
    > Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >
    > Original HD1 and what should be if all OK for HD2
    > Partition 1 = C: WIN9SE (primary) I:
    > Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Partition 3 = F: MASTER (primary) L:
    > Partition 4 = G: WIN2K (primary). M:
    >
    > But, you get C: D:
    > E: & F: G: & H:
    > I: J:
    > ?: ?: (did not look)
    > Ugly results were reported by DOS Fdisk.
    > *
    > Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    > need to use (say) P: as temp placeholder; final result to look like the
    > initial sample (ie: BE IN ORDER on each drive).


    Hey, Robert, I seem to recall that any primary partition has to get
    listed before any extended partitions, so the primary partition on your
    second drive will always be listed before the extended partitions on the
    Primary drive.

    HTH

    Daniel
    , Jan 15, 2013
    #9
  10. Robert Baer

    Robert Baer Guest

    Paul wrote:
    > Robert Baer wrote:
    >> Paul wrote:
    >>> Robert Baer wrote:
    >>>> This very rarely happens to me, but when it does, it is a total (to
    >>>> date) non-recoverable bitch.
    >>>>
    >>>> Background: my HD has the following structure:
    >>>> Partition 1 = C: WIN9SE (primary)
    >>>> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >>>> Partition 3 = F: MASTER (primary)
    >>>> Partition 4 = G: WIN2K (primary).
    >>>> My backup drive is exactly the same.
    >>>>
    >>>> Somehow i did something when both drives were plugged into the
    >>>> computer (i use removable HD kits for all hard drives), and now when i
    >>>> boot with only one HD, i get this incredibly stupid and insane message
    >>>> box "your system has no paging file, or the paging file is too
    >>>> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >>>> procedure of setting the paging file.
    >>>>
    >>>> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >>>> anything other that unplug the computer to turn it off.
    >>>> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >>>> one _then_ boots with both hard drives in the computer.
    >>>>
    >>>> And wasting hours with various page file allocations makes no
    >>>> difference in the above "operation".
    >>>>
    >>>> Please, ANY clues as what else one can do to try and fix this?
    >>>>
    >>>> PS:
    >>>> Do not mention useless "safe mode"; one gets a black screen with "safe
    >>>> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >>>> must unplug computer to reset).
    >>>> This is an artifact when one has a Windoze dual-boot system; i
    >>>> experimented one time without Win98SE (Win2K only OS, boots directly)
    >>>> and sure enough, "safe mode" is actually useable; i damn near fainted
    >>>> due to the shock (many years of useless "safe mode" with a dual-boot
    >>>> system).
    >>>
    >>> I had something like that happen with Win2K.
    >>>
    >>> And the funny part was, even putting a pagefile file onto C:,
    >>> didn't fix it. I even copied a pagefile from another OS,
    >>> thinking that perhaps the pagefile had a header and some info
    >>> that needed to be present, to be recognized.
    >>>
    >>> Which implies some registry info needs to be doctored as well.
    >>> Maybe something like that would have fixed it. (You may be
    >>> able to edit the registry, from another OS.)
    >>>
    >>> I suppose a Repair Install could fix it. As I was in a hurry at the
    >>> time (and was working in a VM), I just started over again, rather
    >>> than waste the time slugging at it.
    >>>
    >>> You could also look at what registry repair options you have
    >>> available. System Restore keeps copies of the registry,
    >>> but I don't think there is System Restore on the older OSes.
    >>>
    >>> Post back if you find a solution.
    >>>
    >>> Paul

    >> "Repair Install" sounds nice, but rarely helps any problem; in this
    >> case it does nothing.
    >> Naturally, there are NO registry repair options that i know of.
    >> And not even Hisself Billy Boy knows where they hide the registry.
    >> RegEdit only works on the registry for that particular OS on its volume.
    >> The only way to fix it is to start from scratch like you mentioned.
    >> And that takes daze to re-build all of the apps.
    >>

    >
    > OK. I fired up my virtual machine and did some testing.
    >
    > 1) I tried a config with just one disk drive, with Win2K C: on it.
    > I deleted C:\pagefile.sys. I restarted. The system created a
    > new pagefile.sys and carried on as if nothing happened. OK,
    > so that's not the recipe on how to break it.
    >
    > 2) Next, I tried two disks with Win2K on them, a C: on each disk.
    > I assigned a label to each disk, so I could tell them apart.
    > I also used Sysinternals VolumeID and gave the second disk a
    > different volume ID. Just in case that was important. (It's possible
    > the OS will do that anyway, but I didn't want to ruin things by
    > ignoring it.)
    >
    > Now, what happens at boot, is interesting. I expected this to
    > happen, with the two disks present.
    >
    > Disk0 C:
    > Disk1 E:
    >
    > And instead, they had switched roles.
    >
    > Disk0 E: Healthy (System)
    > Disk1 C: Healthy (Pagefile)
    >
    > So what's happened is, the system should have booted from Disk0.
    > And it did. But, when it did, two things happened. It decided the
    > other disk had C:\pagefile.sys (which it does). And then it changes the
    > system drive letter to E: ! At this point, no pagefile.sys was
    > deleted.
    >
    > If I delete pagefile.sys from E:, nothing happens, because it's
    > still pointed at Disk1 C:. I can still boot, because both disks
    > are still present, and the pagefile.sys on the so-called C: is
    > still available.
    >
    > OK, now, remove Disk1 and try to run with Disk0. BOom. Now,
    > we're in that pagefile error message loop.
    >
    > And basically, nothing I did would fix it. I tried changing
    > the registry, by editing SYSTEM with a hex editor, and changing
    > that registry entry. (Hmmm. It occurs to me, there are three
    > similar keys, and maybe I edited the wrong one. Mumble...)
    >
    > My conclusion at the moment is, if you were to put back the
    > disks that were present when the "ole switcheroo" was pulled,
    > then the Win2K system would boot. It would look like my mess,
    > with C: having become E: and E: changing to C:. You should put the
    > disks back on the same cables they used originally.
    >
    > So there are actually two variables involved. There is the
    > registry entry that keeps track of system drive letter.
    > And also the registry entry that keeps the pagefile strings.
    > In fact, the pagefile registry entry isn't changing. It's the
    > fact the drive letters have changed, that screws things up.
    > And between the two of those, you have to get things aligned
    > so the system can find the pagefile.
    >
    > Maybe this situation arises, when a person clones Win2K, and
    > both disks with their copy of Win2K remain present, for the very
    > first boot of the clone. Then the drive letter mess happens
    > and so on. And then, disconnecting the second disk, and trying
    > to boot with only one of the two disks present, makes it
    > painfully obvious that the pagefile is missing as far
    > as the system is concerned.
    >
    > Paul

    Because I could easily get Acronis to create that mess, i tried
    changing drive letters, and found it impossible to change drive letters
    on the boot drive, AND on the system drive (that would be C: and E: or
    your sample).
    Now Acronis without any by-your-leave scrambles dive letters which
    virtually guarantees the _possiblity_ of a problem.
    The problem is made real if one leaves both drives connected and
    re-boots.
    The problem is NOT created if one POWERS DOWN the computer at the end
    of the clone process (and removes 2nd drive; i think this step not
    necessary, as re-inserting for a 2-drive state seems to make no problem).
    Robert Baer, Jan 17, 2013
    #10
  11. Robert Baer

    Robert Baer Guest

    Robert Baer wrote:
    > Paul wrote:
    >> Robert Baer wrote:
    >>> Paul wrote:
    >>>> Robert Baer wrote:
    >>>>> This very rarely happens to me, but when it does, it is a total (to
    >>>>> date) non-recoverable bitch.
    >>>>>
    >>>>> Background: my HD has the following structure:
    >>>>> Partition 1 = C: WIN9SE (primary)
    >>>>> Partition 2 = extended with D: WORK (logical) and E: DEVEL (logical)
    >>>>> Partition 3 = F: MASTER (primary)
    >>>>> Partition 4 = G: WIN2K (primary).
    >>>>> My backup drive is exactly the same.
    >>>>>
    >>>>> Somehow i did something when both drives were plugged into the
    >>>>> computer (i use removable HD kits for all hard drives), and now when i
    >>>>> boot with only one HD, i get this incredibly stupid and insane message
    >>>>> box "your system has no paging file, or the paging file is too
    >>>>> small..." and it goes on, ASS-U-MEing one is fully booted, with the
    >>>>> procedure of setting the paging file.
    >>>>>
    >>>>> It is insane and stupid in that it is COMPLETELY IMPOSSIBLE to do
    >>>>> anything other that unplug the computer to turn it off.
    >>>>> There is absolutely NO F*ING WAY to change the paging file(s) unless
    >>>>> one _then_ boots with both hard drives in the computer.
    >>>>>
    >>>>> And wasting hours with various page file allocations makes no
    >>>>> difference in the above "operation".
    >>>>>
    >>>>> Please, ANY clues as what else one can do to try and fix this?
    >>>>>
    >>>>> PS:
    >>>>> Do not mention useless "safe mode"; one gets a black screen with "safe
    >>>>> mode" in all 4 corners and the cursor, and nothing else (oh yes, one
    >>>>> must unplug computer to reset).
    >>>>> This is an artifact when one has a Windoze dual-boot system; i
    >>>>> experimented one time without Win98SE (Win2K only OS, boots directly)
    >>>>> and sure enough, "safe mode" is actually useable; i damn near fainted
    >>>>> due to the shock (many years of useless "safe mode" with a dual-boot
    >>>>> system).
    >>>>
    >>>> I had something like that happen with Win2K.
    >>>>
    >>>> And the funny part was, even putting a pagefile file onto C:,
    >>>> didn't fix it. I even copied a pagefile from another OS,
    >>>> thinking that perhaps the pagefile had a header and some info
    >>>> that needed to be present, to be recognized.
    >>>>
    >>>> Which implies some registry info needs to be doctored as well.
    >>>> Maybe something like that would have fixed it. (You may be
    >>>> able to edit the registry, from another OS.)
    >>>>
    >>>> I suppose a Repair Install could fix it. As I was in a hurry at the
    >>>> time (and was working in a VM), I just started over again, rather
    >>>> than waste the time slugging at it.
    >>>>
    >>>> You could also look at what registry repair options you have
    >>>> available. System Restore keeps copies of the registry,
    >>>> but I don't think there is System Restore on the older OSes.
    >>>>
    >>>> Post back if you find a solution.
    >>>>
    >>>> Paul
    >>> "Repair Install" sounds nice, but rarely helps any problem; in this
    >>> case it does nothing.
    >>> Naturally, there are NO registry repair options that i know of.
    >>> And not even Hisself Billy Boy knows where they hide the registry.
    >>> RegEdit only works on the registry for that particular OS on its volume.
    >>> The only way to fix it is to start from scratch like you mentioned.
    >>> And that takes daze to re-build all of the apps.
    >>>

    >>
    >> OK. I fired up my virtual machine and did some testing.
    >>
    >> 1) I tried a config with just one disk drive, with Win2K C: on it.
    >> I deleted C:\pagefile.sys. I restarted. The system created a
    >> new pagefile.sys and carried on as if nothing happened. OK,
    >> so that's not the recipe on how to break it.
    >>
    >> 2) Next, I tried two disks with Win2K on them, a C: on each disk.
    >> I assigned a label to each disk, so I could tell them apart.
    >> I also used Sysinternals VolumeID and gave the second disk a
    >> different volume ID. Just in case that was important. (It's possible
    >> the OS will do that anyway, but I didn't want to ruin things by
    >> ignoring it.)
    >>
    >> Now, what happens at boot, is interesting. I expected this to
    >> happen, with the two disks present.
    >>
    >> Disk0 C:
    >> Disk1 E:
    >>
    >> And instead, they had switched roles.
    >>
    >> Disk0 E: Healthy (System)
    >> Disk1 C: Healthy (Pagefile)
    >>
    >> So what's happened is, the system should have booted from Disk0.
    >> And it did. But, when it did, two things happened. It decided the
    >> other disk had C:\pagefile.sys (which it does). And then it changes the
    >> system drive letter to E: ! At this point, no pagefile.sys was
    >> deleted.
    >>
    >> If I delete pagefile.sys from E:, nothing happens, because it's
    >> still pointed at Disk1 C:. I can still boot, because both disks
    >> are still present, and the pagefile.sys on the so-called C: is
    >> still available.
    >>
    >> OK, now, remove Disk1 and try to run with Disk0. BOom. Now,
    >> we're in that pagefile error message loop.
    >>
    >> And basically, nothing I did would fix it. I tried changing
    >> the registry, by editing SYSTEM with a hex editor, and changing
    >> that registry entry. (Hmmm. It occurs to me, there are three
    >> similar keys, and maybe I edited the wrong one. Mumble...)
    >>
    >> My conclusion at the moment is, if you were to put back the
    >> disks that were present when the "ole switcheroo" was pulled,
    >> then the Win2K system would boot. It would look like my mess,
    >> with C: having become E: and E: changing to C:. You should put the
    >> disks back on the same cables they used originally.
    >>
    >> So there are actually two variables involved. There is the
    >> registry entry that keeps track of system drive letter.
    >> And also the registry entry that keeps the pagefile strings.
    >> In fact, the pagefile registry entry isn't changing. It's the
    >> fact the drive letters have changed, that screws things up.
    >> And between the two of those, you have to get things aligned
    >> so the system can find the pagefile.
    >>
    >> Maybe this situation arises, when a person clones Win2K, and
    >> both disks with their copy of Win2K remain present, for the very
    >> first boot of the clone. Then the drive letter mess happens
    >> and so on. And then, disconnecting the second disk, and trying
    >> to boot with only one of the two disks present, makes it
    >> painfully obvious that the pagefile is missing as far
    >> as the system is concerned.
    >>
    >> Paul

    > I think i have found a way to reliably create the problem as well as fix
    > it.
    > Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >
    > Original HD1 and what should be if all OK for HD2
    > Partition 1 = C: WIN9SE (primary) I:
    > Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Partition 3 = F: MASTER (primary) L:
    > Partition 4 = G: WIN2K (primary). M:
    >
    > But, you get C: D:
    > E: & F: G: & H:
    > I: J:
    > ?: ?: (did not look)
    > Ugly results were reported by DOS Fdisk.
    > *
    > Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    > need to use (say) P: as temp placeholder; final result to look like the
    > initial sample (ie: BE IN ORDER on each drive).

    Found: CANNOT be fixed.
    Robert Baer, Jan 17, 2013
    #11
  12. Robert Baer

    Robert Baer Guest

    wrote:
    > Robert Baer wrote:
    >
    > <Snip>
    >>> Paul

    >> I think i have found a way to reliably create the problem as well as
    >> fix it.
    >> Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >>
    >> Original HD1 and what should be if all OK for HD2
    >> Partition 1 = C: WIN9SE (primary) I:
    >> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >> Partition 3 = F: MASTER (primary) L:
    >> Partition 4 = G: WIN2K (primary). M:
    >>
    >> But, you get C: D:
    >> E: & F: G: & H:
    >> I: J:
    >> ?: ?: (did not look)
    >> Ugly results were reported by DOS Fdisk.
    >> *
    >> Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    >> need to use (say) P: as temp placeholder; final result to look like the
    >> initial sample (ie: BE IN ORDER on each drive).

    >
    > Hey, Robert, I seem to recall that any primary partition has to get
    > listed before any extended partitions, so the primary partition on your
    > second drive will always be listed before the extended partitions on the
    > Primary drive.
    >
    > HTH
    >
    > Daniel
    >

    Well, on all drives i have, whether bad "clones" (read: made by
    Acronis) or not, the layout is Primary, Extended, Primary, Primary.
    So..whatever you had in mind, may not be in effect.
    Robert Baer, Jan 17, 2013
    #12
  13. Robert Baer

    Guest

    Robert Baer wrote:
    > wrote:
    >> Robert Baer wrote:
    >>
    >> <Snip>
    >>>> Paul
    >>> I think i have found a way to reliably create the problem as well as
    >>> fix it.
    >>> Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >>>
    >>> Original HD1 and what should be if all OK for HD2
    >>> Partition 1 = C: WIN9SE (primary) I:
    >>> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >>> Partition 3 = F: MASTER (primary) L:
    >>> Partition 4 = G: WIN2K (primary). M:
    >>>
    >>> But, you get C: D:
    >>> E: & F: G: & H:
    >>> I: J:
    >>> ?: ?: (did not look)
    >>> Ugly results were reported by DOS Fdisk.
    >>> *
    >>> Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    >>> need to use (say) P: as temp placeholder; final result to look like the
    >>> initial sample (ie: BE IN ORDER on each drive).

    >>
    >> Hey, Robert, I seem to recall that any primary partition has to get
    >> listed before any extended partitions, so the primary partition on your
    >> second drive will always be listed before the extended partitions on the
    >> Primary drive.
    >>
    >> HTH
    >>
    >> Daniel
    >>

    > Well, on all drives i have, whether bad "clones" (read: made by
    > Acronis) or not, the layout is Primary, Extended, Primary, Primary.
    > So..whatever you had in mind, may not be in effect.


    Robert, lets say, on your main drive, you have:-
    Partition 1 = C: WIN9SE (primary) I:
    Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    Partition 3 = F: MASTER (primary) L:
    Partition 4 = G: WIN2K (primary). M:

    When you clone this set-up to another drive, it will also have:-
    Partition 1 = C: WIN9SE (primary) I:
    Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    Partition 3 = F: MASTER (primary) L:
    Partition 4 = G: WIN2K (primary). M:

    But, so that we can differentiate between the two drives, lets call this
    second lot:-
    Clone Partition 1 = C: WIN9SE (primary) I:
    Clone Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    Clone Partition 3 = F: MASTER (primary) L:
    Clone Partition 4 = G: WIN2K (primary). M:

    What you are the seeing, in Windows Explorer (or whatever), with the two
    hard drives still connected, would be something like:-
    Partition 1 = C: WIN9SE (primary) I:
    Clone Partition 1 = C: WIN9SE (primary) I:
    Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    Clone Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    Partition 3 = F: MASTER (primary) L:
    Partition 4 = G: WIN2K (primary). M:
    Clone Partition 3 = F: MASTER (primary) L:
    Clone Partition 4 = G: WIN2K (primary). M:

    Note the two C:\ drives would appear first, then at least one of your
    "D:\, E:\, J:\ and K:\" drives, then (I'm guessing) the other "D:\, E:\,
    J:\ and K:\" drive then the two F:\ and G:\ drive partitions from your
    main drive, then the two F:\ and G:\ drive partitions from your clone drive.

    Although a possibility might be the two C:\ drives would appear first,
    then at least one of your "D:\, E:\, J:\ and K:\" drives, then the F:\
    and G:\ drive partitions from your main drive, then (maybe) the other
    "D:\, E:\, J:\ and K:\" drive then the F:\ and G:\ drive partitions from
    your clone drive.

    Something like that....I think.....maybe!!

    Daniel
    , Jan 17, 2013
    #13
  14. Robert Baer

    Robert Baer Guest

    wrote:
    > Robert Baer wrote:
    >> wrote:
    >>> Robert Baer wrote:
    >>>
    >>> <Snip>
    >>>>> Paul
    >>>> I think i have found a way to reliably create the problem as well as
    >>>> fix it.
    >>>> Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >>>>
    >>>> Original HD1 and what should be if all OK for HD2
    >>>> Partition 1 = C: WIN9SE (primary) I:
    >>>> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >>>> Partition 3 = F: MASTER (primary) L:
    >>>> Partition 4 = G: WIN2K (primary). M:
    >>>>
    >>>> But, you get C: D:
    >>>> E: & F: G: & H:
    >>>> I: J:
    >>>> ?: ?: (did not look)
    >>>> Ugly results were reported by DOS Fdisk.
    >>>> *
    >>>> Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    >>>> need to use (say) P: as temp placeholder; final result to look like the
    >>>> initial sample (ie: BE IN ORDER on each drive).
    >>>
    >>> Hey, Robert, I seem to recall that any primary partition has to get
    >>> listed before any extended partitions, so the primary partition on your
    >>> second drive will always be listed before the extended partitions on the
    >>> Primary drive.
    >>>
    >>> HTH
    >>>
    >>> Daniel
    >>>

    >> Well, on all drives i have, whether bad "clones" (read: made by
    >> Acronis) or not, the layout is Primary, Extended, Primary, Primary.
    >> So..whatever you had in mind, may not be in effect.

    >
    > Robert, lets say, on your main drive, you have:-
    > Partition 1 = C: WIN9SE (primary) I:
    > Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Partition 3 = F: MASTER (primary) L:
    > Partition 4 = G: WIN2K (primary). M:

    * If that was the only HD, or one was using GhostPE, the above drive
    designations would be correct. CD/DVD drive designations come next (say
    one for your example below, H:).

    >
    > When you clone this set-up to another drive, it will also have:-
    > Partition 1 = C: WIN9SE (primary) I:
    > Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Partition 3 = F: MASTER (primary) L:
    > Partition 4 = G: WIN2K (primary). M:

    * That is exactly the case for GhostPE. BUT Acronis scrambles the
    designations to something like this:
    Partition 1 2 3 4
    HD1 C E, F J L
    HD2 D G, I K M.
    That means, for an original HD1 swapfile designation of G:, one can
    get royally screwed with a useless and non-recoverable drive.
    If one is aware of the consequences, and does exactly the correct
    procedure, the result can be avoided; namely SHUT DOWN and POWER OFF the
    computer _immediately_ after the cloning is completed.

    >
    > But, so that we can differentiate between the two drives, lets call this
    > second lot:-
    > Clone Partition 1 = C: WIN9SE (primary) I:
    > Clone Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Clone Partition 3 = F: MASTER (primary) L:
    > Clone Partition 4 = G: WIN2K (primary). M:
    >
    > What you are the seeing, in Windows Explorer (or whatever), with the two
    > hard drives still connected, would be something like:-
    > Partition 1 = C: WIN9SE (primary) I:
    > Clone Partition 1 = C: WIN9SE (primary) I:
    > Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Clone Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    > Partition 3 = F: MASTER (primary) L:
    > Partition 4 = G: WIN2K (primary). M:
    > Clone Partition 3 = F: MASTER (primary) L:
    > Clone Partition 4 = G: WIN2K (primary). M:
    >
    > Note the two C:\ drives would appear first, then at least one of your
    > "D:\, E:\, J:\ and K:\" drives, then (I'm guessing) the other "D:\, E:\,
    > J:\ and K:\" drive then the two F:\ and G:\ drive partitions from your
    > main drive, then the two F:\ and G:\ drive partitions from your clone
    > drive.
    >
    > Although a possibility might be the two C:\ drives would appear first,
    > then at least one of your "D:\, E:\, J:\ and K:\" drives, then the F:\
    > and G:\ drive partitions from your main drive, then (maybe) the other
    > "D:\, E:\, J:\ and K:\" drive then the F:\ and G:\ drive partitions from
    > your clone drive.
    >
    > Something like that....I think.....maybe!!
    >
    > Daniel
    >

    * Nice in theory; works for GhostPE, totally AFU with Acronis as
    mentioned above.
    When on selects the drives for setting up the cloning, the "preview"
    _shows_ the scramble.
    With GhostPE, drive letters are in a linear and comprehensive order;
    NOT interleaved.

    Naturally, that is not the only serious problem with Acronis;
    partition placement, lengths and endpoints get messed, along with the
    one-byte type marker getting altered (hex06 changed to hex0E).
    FAT entries also get altered.

    Once upon a time, one used to use the dictionary and agree with the
    meaning of certain words.
    Robert Baer, Jan 18, 2013
    #14
  15. Robert Baer

    Guest

    Robert Baer wrote:
    > wrote:
    >> Robert Baer wrote:
    >>> wrote:
    >>>> Robert Baer wrote:
    >>>>
    >>>> <Snip>
    >>>>>> Paul
    >>>>> I think i have found a way to reliably create the problem as well as
    >>>>> fix it.
    >>>>> Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >>>>>
    >>>>> Original HD1 and what should be if all OK for HD2
    >>>>> Partition 1 = C: WIN9SE (primary) I:
    >>>>> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >>>>> Partition 3 = F: MASTER (primary) L:
    >>>>> Partition 4 = G: WIN2K (primary). M:
    >>>>>
    >>>>> But, you get C: D:
    >>>>> E: & F: G: & H:
    >>>>> I: J:
    >>>>> ?: ?: (did not look)
    >>>>> Ugly results were reported by DOS Fdisk.
    >>>>> *
    >>>>> Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    >>>>> need to use (say) P: as temp placeholder; final result to look like
    >>>>> the
    >>>>> initial sample (ie: BE IN ORDER on each drive).
    >>>>
    >>>> Hey, Robert, I seem to recall that any primary partition has to get
    >>>> listed before any extended partitions, so the primary partition on your
    >>>> second drive will always be listed before the extended partitions on
    >>>> the
    >>>> Primary drive.
    >>>>
    >>>> HTH
    >>>>
    >>>> Daniel
    >>>>
    >>> Well, on all drives i have, whether bad "clones" (read: made by
    >>> Acronis) or not, the layout is Primary, Extended, Primary, Primary.
    >>> So..whatever you had in mind, may not be in effect.

    >>
    >> Robert, lets say, on your main drive, you have:-
    >> Partition 1 = C: WIN9SE (primary) I:
    >> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >> Partition 3 = F: MASTER (primary) L:
    >> Partition 4 = G: WIN2K (primary). M:

    > * If that was the only HD, or one was using GhostPE, the above drive
    > designations would be correct. CD/DVD drive designations come next (say
    > one for your example below, H:).
    >
    >>
    >> When you clone this set-up to another drive, it will also have:-
    >> Partition 1 = C: WIN9SE (primary) I:
    >> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >> Partition 3 = F: MASTER (primary) L:
    >> Partition 4 = G: WIN2K (primary). M:

    > * That is exactly the case for GhostPE. BUT Acronis scrambles the
    > designations to something like this:
    > Partition 1 2 3 4
    > HD1 C E, F J L
    > HD2 D G, I K M.
    > That means, for an original HD1 swapfile designation of G:, one can
    > get royally screwed with a useless and non-recoverable drive.
    > If one is aware of the consequences, and does exactly the correct
    > procedure, the result can be avoided; namely SHUT DOWN and POWER OFF the
    > computer _immediately_ after the cloning is completed.


    Robert, have you tried removing/disconnecting your Hard Drive (the one
    that contains your current C:\, D:\, E:\, F:\, and G:\ drives) and
    install the Hard Disk onto which you have just backed everything up on to?

    You know, just to check that your Back-up works!!

    Daniel
    , Jan 20, 2013
    #15
  16. Robert Baer

    Robert Baer Guest

    wrote:
    > Robert Baer wrote:
    >> wrote:
    >>> Robert Baer wrote:
    >>>> wrote:
    >>>>> Robert Baer wrote:
    >>>>>
    >>>>> <Snip>
    >>>>>>> Paul
    >>>>>> I think i have found a way to reliably create the problem as well as
    >>>>>> fix it.
    >>>>>> Use Acronis to clone HD1 to HD2 and re-boot with both connected.
    >>>>>>
    >>>>>> Original HD1 and what should be if all OK for HD2
    >>>>>> Partition 1 = C: WIN9SE (primary) I:
    >>>>>> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >>>>>> Partition 3 = F: MASTER (primary) L:
    >>>>>> Partition 4 = G: WIN2K (primary). M:
    >>>>>>
    >>>>>> But, you get C: D:
    >>>>>> E: & F: G: & H:
    >>>>>> I: J:
    >>>>>> ?: ?: (did not look)
    >>>>>> Ugly results were reported by DOS Fdisk.
    >>>>>> *
    >>>>>> Fix: Boot to Win2K and use disk manager to re-assign the drives (will
    >>>>>> need to use (say) P: as temp placeholder; final result to look like
    >>>>>> the
    >>>>>> initial sample (ie: BE IN ORDER on each drive).
    >>>>>
    >>>>> Hey, Robert, I seem to recall that any primary partition has to get
    >>>>> listed before any extended partitions, so the primary partition on
    >>>>> your
    >>>>> second drive will always be listed before the extended partitions on
    >>>>> the
    >>>>> Primary drive.
    >>>>>
    >>>>> HTH
    >>>>>
    >>>>> Daniel
    >>>>>
    >>>> Well, on all drives i have, whether bad "clones" (read: made by
    >>>> Acronis) or not, the layout is Primary, Extended, Primary, Primary.
    >>>> So..whatever you had in mind, may not be in effect.
    >>>
    >>> Robert, lets say, on your main drive, you have:-
    >>> Partition 1 = C: WIN9SE (primary) I:
    >>> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >>> Partition 3 = F: MASTER (primary) L:
    >>> Partition 4 = G: WIN2K (primary). M:

    >> * If that was the only HD, or one was using GhostPE, the above drive
    >> designations would be correct. CD/DVD drive designations come next (say
    >> one for your example below, H:).
    >>
    >>>
    >>> When you clone this set-up to another drive, it will also have:-
    >>> Partition 1 = C: WIN9SE (primary) I:
    >>> Partition 2 = extended D: WORK & E: DEVEL J: & K: (logical)
    >>> Partition 3 = F: MASTER (primary) L:
    >>> Partition 4 = G: WIN2K (primary). M:

    >> * That is exactly the case for GhostPE. BUT Acronis scrambles the
    >> designations to something like this:
    >> Partition 1 2 3 4
    >> HD1 C E, F J L
    >> HD2 D G, I K M.
    >> That means, for an original HD1 swapfile designation of G:, one can
    >> get royally screwed with a useless and non-recoverable drive.
    >> If one is aware of the consequences, and does exactly the correct
    >> procedure, the result can be avoided; namely SHUT DOWN and POWER OFF the
    >> computer _immediately_ after the cloning is completed.

    >
    > Robert, have you tried removing/disconnecting your Hard Drive (the one
    > that contains your current C:\, D:\, E:\, F:\, and G:\ drives) and
    > install the Hard Disk onto which you have just backed everything up on to?
    >
    > You know, just to check that your Back-up works!!
    >
    > Daniel
    >

    The clone that Acronis makes cannot work as-is.
    First, one must use PTedit (from Partition Magic) to edit the
    partition type byte for partition #3 - else nothing can see it or get to it.
    Then, one needs to edit start cylinder, end cylinder, and length.
    Finally, must shut down and re-start.

    Once upon a time, "copy" (a 4-letter word) had a reliable and useful
    meaning; ditto for "clone".
    Robert Baer, Jan 21, 2013
    #16
  17. Robert Baer

    Guest

    Robert Baer wrote:

    <Snip>
    > Once upon a time, "copy" (a 4-letter word) had a reliable and useful
    > meaning; ditto for "clone".


    Yeap!! Back in "the good old days"!!

    Daniel
    , Jan 22, 2013
    #17
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Phil
    Replies:
    7
    Views:
    1,055
    Wyatt M. Portendt
    Mar 4, 2004
  2. Fred-C-
    Replies:
    1
    Views:
    356
  3. TrickTrash
    Replies:
    12
    Views:
    3,781
    TrickTrash
    Jan 21, 2007
  4. Cindy Stuart

    Help - browser loads another page after home page

    Cindy Stuart, Apr 28, 2007, in forum: Computer Support
    Replies:
    4
    Views:
    541
    Cindy Stuart
    Apr 29, 2007
  5. Jon Mitchell

    Page file on 64 Bit System

    Jon Mitchell, Nov 19, 2009, in forum: Windows 64bit
    Replies:
    3
    Views:
    7,053
    bluefin
    Nov 23, 2009
Loading...

Share This Page