D: Drive Disappeared

[Bob]

For no apparent reason, I lost the D: drive in Win2K. It shows up on
the BIOS POST, so I know it's there. The IDE channel is present in
Device Manager. Devide Manager also reports that the physical drive is
present and working properly. It just does not show up in Win2K.

What is going on and what do I need to do to get it back online?

 

[kony]

For no apparent reason, I lost the D: drive in Win2K. It shows up on
the BIOS POST, so I know it's there. The IDE channel is present in
Device Manager. Devide Manager also reports that the physical drive is
present and working properly. It just does not show up in Win2K.

What is going on and what do I need to do to get it back online?

What do Disk Management and the Event Viewer show?

If it's Fat32, try booting DOS to see if it's visible. If
NTFS, Google for your preferred method of reading NTFS
outside of windows. IIRC, http://www.sysinternals.com has a
NTFS filesystem driver for DOS, or use something else like a
Bart's PE disc.

 

[Bob]

What do Disk Management and the Event Viewer show?

Disk Management does not show the drive.

Event Viewer has no messages relating to this problem.

If it's Fat32, try booting DOS to see if it's visible.

It's NTFS.

If >NTFS, Google for your preferred method of reading NTFS
outside of windows. IIRC, http://www.sysinternals.com has a
NTFS filesystem driver for DOS, or use something else like a
Bart's PE disc.

I already have NTFS4DOS from Datapol.

I used it with the drives in their usual position, but could not find
D: (I did find C. Then I put the D: drive in the C: bay and I could
see it with NTFS4DOS. Weird.

That made me think the D: removable bay was the problem so I put a
different tray in that bay. I booted to Win2K and saw both drives.
Weird.

So I put the D: drive in its bay and booted Win2K and this time D:
showed up. I rebooted a couple more times to make sure and every time
D: showed up. Weird.

I ran WD DataLifeGuard and it said the drive was OK. I ran Win98 FDISK
/MBR in case the MBR was corrupted, and it ran OK but put out a
message that the MBR was not updated. Weird.

Maybe one or more of those maneuvers on my part restored something
that allowed the drive to be recognized. For example maybe WD DLG did
something to the MBR and that fixed it.

I hate it when things like this happen. There is no excuse for Win2K
not to tell me what is wrong instead of sitting there like a castrated
imbecile.

The only guess is something may have corrupted the MBR. Would that
cause the problem? Remember that I saw the drive in Device Manager -
it said it was OK.

 

[~misfit~]

Disk Management does not show the drive.

Event Viewer has no messages relating to this problem.


It's NTFS.


I already have NTFS4DOS from Datapol.

I used it with the drives in their usual position, but could not find
D: (I did find C. Then I put the D: drive in the C: bay and I could
see it with NTFS4DOS. Weird.

That made me think the D: removable bay was the problem so I put a
different tray in that bay. I booted to Win2K and saw both drives.
Weird.

So I put the D: drive in its bay and booted Win2K and this time D:
showed up. I rebooted a couple more times to make sure and every time
D: showed up. Weird.

I ran WD DataLifeGuard and it said the drive was OK. I ran Win98 FDISK
/MBR in case the MBR was corrupted, and it ran OK but put out a
message that the MBR was not updated. Weird.

Because 98 won't update an NTFS-formatted MBR.

Maybe one or more of those maneuvers on my part restored something
that allowed the drive to be recognized. For example maybe WD DLG did
something to the MBR and that fixed it.

I hate it when things like this happen. There is no excuse for Win2K
not to tell me what is wrong instead of sitting there like a castrated
imbecile.

The only guess is something may have corrupted the MBR. Would that
cause the problem? Remember that I saw the drive in Device Manager -
it said it was OK.

I think that the problem is rather obvious. These 'removable bays' that
you're using are causing problems. Intermittant problems as sometimes the
drive shows up. You changing the trays around cleaned off the gunk/oxidation
on the connectors a bit and that's why it's working OK now. I suggest you
clean the contacts at the back of the bay and the tray thouroughly with CO
cleaner, meths or something similar.

How do you expect W2K to say "It's time to clean the contacts on your
removable HDD trays"? Hell you didn't even tell us you were using them in
your first post, have you told W2K?

 

[kony]

Disk Management does not show the drive.

Event Viewer has no messages relating to this problem.


It's NTFS.


I already have NTFS4DOS from Datapol.

I used it with the drives in their usual position, but could not find
D: (I did find C. Then I put the D: drive in the C: bay and I could
see it with NTFS4DOS. Weird.

That made me think the D: removable bay was the problem so I put a
different tray in that bay. I booted to Win2K and saw both drives.
Weird.

So I put the D: drive in its bay and booted Win2K and this time D:
showed up. I rebooted a couple more times to make sure and every time
D: showed up. Weird.

Your removable bays might have intermittent connections. As
a matter of routine you might pull the drive out and plug it
back in every so often, but on the other hand it could even
exacerbate the problem. I'd probably keep the drive out of
the removable bay till this gets sorted out.

Maybe one or more of those maneuvers on my part restored something
that allowed the drive to be recognized. For example maybe WD DLG did
something to the MBR and that fixed it.

I hate it when things like this happen. There is no excuse for Win2K
not to tell me what is wrong instead of sitting there like a castrated
imbecile.

Well how could it know? If it's not a software problem and
the drive isn't there to "tell" it what's going on, there's
nothing any OS could do except guess.

The only guess is something may have corrupted the MBR. Would that
cause the problem? Remember that I saw the drive in Device Manager -
it said it was OK.

I'd suspect the removable bay.

 

[Bob]

Because 98 won't update an NTFS-formatted MBR.

Hmm... It works on my other NTFS-formatted disks.

I use it to bugger the disk signature when I make clones. Win2K tries
to defeat clone making by keeping signatures for each disk and BSODing
when it finds two disks with the same signature. Win98 FDISK has a bug
that causes it to put zeros in the first 4 bytes of the signature,
thereby changing the signature so you can at least mount the clone
with the original. I have used that "feature" many times with success.
Therefore Win98 FDISK must be doing something right. Also it does not
issue any error messages like it did with the disk that Win2K would
not recognize.

I think that the problem is rather obvious. These 'removable bays' that
you're using are causing problems. Intermittant problems as sometimes the
drive shows up.

I have no reason to believe that, and even if it were true it is far
from obvious.

You changing the trays around cleaned off the gunk/oxidation
on the connectors a bit and that's why it's working OK now. I suggest you
clean the contacts at the back of the bay and the tray thouroughly with CO
cleaner, meths or something similar.

I use the bays often enough to keep them clean without having to
resort to cleaning them manually.

 

[Bob]

Your removable bays might have intermittent connections. As
a matter of routine you might pull the drive out and plug it
back in every so often, but on the other hand it could even
exacerbate the problem. I'd probably keep the drive out of
the removable bay till this gets sorted out.

I have no reason to believe the removable bays are the problem. When I
put a good disk in the lower bay it showed up in Win2K.

You are forgetting that when the D: was not recognized the Device
Manager showed the disk present and the IDE channel present. The only
thing was that the disk was not recognizes, which leads me to believe
something on the disk was corrupt, like the MBR.

Well how could it know? If it's not a software problem and
the drive isn't there to "tell" it what's going on, there's
nothing any OS could do except guess.

UNIX would give the user enough information to diagnose the problem.
Win2K just gives you a blank stare.

BTW, I tried to run the "troubleshooting" but it came back with some
cryptic nonsense about how it can't run scripts. It then asked me if I
wanted to run scripts or not, and no matter which I chose, it put up a
"not found" page similar to IE when it can't find a web page.

I'd suspect the removable bay.

It could be the tray. If it were the bay itself then I would not have
seen the other disk I used to test that possibility.

 

[~misfit~]

I have no reason to believe the removable bays are the problem. When I
put a good disk in the lower bay it showed up in Win2K.

You are forgetting that when the D: was not recognized the Device
Manager showed the disk present and the IDE channel present. The only
thing was that the disk was not recognizes, which leads me to believe
something on the disk was corrupt, like the MBR.


UNIX would give the user enough information to diagnose the problem.
Win2K just gives you a blank stare.

BTW, I tried to run the "troubleshooting" but it came back with some
cryptic nonsense about how it can't run scripts. It then asked me if I
wanted to run scripts or not, and no matter which I chose, it put up a
"not found" page similar to IE when it can't find a web page.


It could be the tray. If it were the bay itself then I would not have
seen the other disk I used to test that possibility.

Unless the act of removing one tray and inserting another "fixed" it.

 

[~misfit~]

Hmm... It works on my other NTFS-formatted disks.

I use it to bugger the disk signature when I make clones. Win2K tries
to defeat clone making by keeping signatures for each disk and BSODing
when it finds two disks with the same signature. Win98 FDISK has a bug
that causes it to put zeros in the first 4 bytes of the signature,
thereby changing the signature so you can at least mount the clone
with the original. I have used that "feature" many times with success.
Therefore Win98 FDISK must be doing something right. Also it does not
issue any error messages like it did with the disk that Win2K would
not recognize.

I'll take your word for it, I'm a (mainly) hardware man.

I have no reason to believe that, and even if it were true it is far
from obvious.

To you maybe.

I use the bays often enough to keep them clean without having to
resort to cleaning them manually.

And you don't think that maybe because you use them "often enough" that
there could perhaps be another problem like worn contacts?

My money's still on the tray/bay thing. It's the best explaination, going by
what information you've provided. I've been troubleshooting other people's
PCs for years now and you get a 'feel' for things. I just fixed a problem
that had someone stumped all day in 30 seconds tonight. There's no
substitute for hands-on experience.

 

[kony]

I have no reason to believe the removable bays are the problem. When I
put a good disk in the lower bay it showed up in Win2K.

This very issue IS a reason to suspect the removable bays.
Pulling one out and putting it in the same or even a
different bay slot and then having them work is no sign
they're ok.

Ever seen a dodgy AC electrical outlet where you had to
wiggle the plug to get contact? Could be similar scenario
except that contact had to be good and stay good before
system is booted, you can't then later wiggle it and have a
good logical state of the device with many IDE controllers.

You are forgetting that when the D: was not recognized the Device
Manager showed the disk present and the IDE channel present. The only
thing was that the disk was not recognizes, which leads me to believe
something on the disk was corrupt, like the MBR.

Are the files still there? If so, your partition table is
intact. Given that, and that you're not trying to boot it,
what would it matter about the MBR?

UNIX would give the user enough information to diagnose the problem.
Win2K just gives you a blank stare.

So you have been running unix on that box and have been
provided such a diagnostic report? Otherwise it's just
speculation without even knowing the problem yet.

BTW, I tried to run the "troubleshooting" but it came back with some
cryptic nonsense about how it can't run scripts. It then asked me if I
wanted to run scripts or not, and no matter which I chose, it put up a
"not found" page similar to IE when it can't find a web page.


It could be the tray. If it were the bay itself then I would not have
seen the other disk I used to test that possibility.

Tray, bay, somewhat the same thing, aren't they supposed to
be a working set?

I doubt "troubleshooting" would help here, an OS tends to
troubleshoot itself, not a physical device not working.
Even so, you probably have some kind of scripting support
disabled either by yourself or some 3rd party security
related software. Offhand I don't know what but look around
on the system for what might've changed windows' settings.

 

[Bob]

To you maybe.

What is that supposed to mean - that you know something I don't know?
What might that be?

I tried the lower bay with another disk and it worked. Win2K said the
missing disk was present at the hardware level and working. The
problem was at the logical drive level, not hardware level.

My money's still on the tray/bay thing. It's the best explaination, going by
what information you've provided. I've been troubleshooting other people's
PCs for years now and you get a 'feel' for things. I just fixed a problem
that had someone stumped all day in 30 seconds tonight. There's no
substitute for hands-on experience.

And what makes you think you are the only person who has enough
experience? How many patents do you have for electronic instrument
design?

If it were the tray/bay thing that was bad, then why did the disk show
up in Win2K Device Manager as working properly?

When I ran Win98 FDISK /MBR, the disk was present or else FDISK would
have said "device missing" or somesuch. Instead it reported an MBR
problem, which would be impossible if the tray/bay thing were
malfunctioning.

Do you always jump to the wrong conclusions when you do not consider
all the facts?

 

[Bob]

Unless the act of removing one tray and inserting another "fixed" it.

Before I swapped trays, Win2K Device Manager reported that the device
was present and working properly. That's a hardware consideration. The
problem is in the software layer - the logical layer - where the
device is mounted. Something must have corrupted the MBR which
prevented the disk from being mounted. Running WD DLG diags fixed the
MBR.

 

[Bob]

This very issue IS a reason to suspect the removable bays.

Please read the previous posts to see why I rule that out.

If the "tray/bay thing" as it has been designated by a self-proclaimed
hardware expert, were the culprit, then Win2K Device Manager would not
have said that the device was present and functioning properly.

I believe the problem was in a corrupt MBR which WD DLG diags
repaired. Recall that I have had problems with the C: drive becoming
corrupted so maybe it was the D: drive's turn.

I have seen this very thing before when I try to mount two identical
drives, one a clone of the other. In that case the symptoms are
identical - the device shows up but will not mount. That is a logical
device problem, not a hardware problem.

Hopefully this will be the end of it for a while. I just learned that
Vista is going to be delayed so there went my hopes of rebuilding the
operating system for the foreseeable future.

I may have set some kind of record in terms of not rebuilding Windows.
I started out with NT4 in 1997 and went thru all the SP and then
loaded Win2K in 2000 as an Upgrade over NT4. I have never actually
installed Win2K from scratch. I have done an "in place upgrade" a
couple of times and maybe I should do it again but it is such a bitch
sorting out the things that it screws up.

I keep telling myself that I will not reinstall until Vista because
it's just around the corner. HA!

 

[kony]

What is that supposed to mean - that you know something I don't know?
What might that be?

I'm sure he does know something you don't. I know something
you don't too, and you know something we don't. Either
we're all set to play Trivial Pursuit or there can be other
gains.

I tried the lower bay with another disk and it worked. Win2K said the
missing disk was present at the hardware level and working.

Ok but, what if you keep using that other drive in this
situation? Presumably the drive now having trouble, also
seemed to work ok initially?

The
problem was at the logical drive level, not hardware level.

You might be right but it's a premature conclusion to make,
and one that if wrong, could result in a lot of unfruitfull
effort.

And what makes you think you are the only person who has enough
experience? How many patents do you have for electronic instrument
design?

LOL.

Ok Mr Instrument Designer, would you want your removeable
drive enclosure to house the entire interface between the
fly-by-wire and the pilot on the next plane you fly on?

They're not terribly precision devices, these drive bays.
The connectors alone would cost more than the entire
enclosure if that were the case.

If it were the tray/bay thing that was bad, then why did the disk show
up in Win2K Device Manager as working properly?

Because at that moment, it was? Such is the nature of
intermittent connections, gone one minute and back the next.
If you had some stray data that wasn't written that could be
a big problem. Now if the drive is corrupt as a result but
momentarily connected, you'd see it.

When I ran Win98 FDISK /MBR, the disk was present or else FDISK would
have said "device missing" or somesuch. Instead it reported an MBR
problem, which would be impossible if the tray/bay thing were
malfunctioning.

No, it would only be impossible for it to see the drive if
it were completely unworkable. Obviously even a dodgy
design has to be at least a little workable or it'd never
sell, and/or RMA would eat the company alive.

Do you always jump to the wrong conclusions when you do not consider
all the facts?

If we had ALL the facts, would the problem still persist?

IMO, it's likely one of 4 things:

- Removable drive bay, JUST REMOVE IT!
Get it out of the equation because the only sane way to
tackle your problem is elimination of variables.

- Research your motherboard and drive controller... is it
subject to any bugs like this?

- Drive itself, maybe bad. Are you sure you have it
jumpered right? Don't forget that WD drives need a
different Single vs. Master w/slave jumper setting, changed
when put in a different logical arrangement (Possible with
different bays).

- OS corrupt in some way or 3rd party has changed it. You
could hunt and comb the registry, research the fast
shutdown cache flushing issues or just do it the easy way:
Slipstream SP4 into Win2k and do a clean install. Not
"recreate your entire working environment", just a testbed
to see if you still have the problem. That is, after you
get the drive working again, logicall intact
data/MBR/partitions/etc.

- When all else fails, do what everyone else eventually
moves towards, not putting anything important on a "PC"
system but rather networked into a NAS of one sort or
another (preferribly not one of those tiny ez-bake-oven
types that overheats drives too easily but rather a
corporate type drive enclosure with it's own switcher and
fan inside or a random old system turned fileserver).

As for not removing the removable drive bays entirely, you
are nuts to have not done this already given the problems
you've been having retaining data. What's the point of
being able to remove an empty drive???

 

[kony]

Please read the previous posts to see why I rule that out.

If the "tray/bay thing" as it has been designated by a self-proclaimed
hardware expert, were the culprit, then Win2K Device Manager would not
have said that the device was present and functioning properly.

not necessarily, think "intermittent connection".

I believe the problem was in a corrupt MBR which WD DLG diags
repaired. Recall that I have had problems with the C: drive becoming
corrupted so maybe it was the D: drive's turn.

Even if it were corrupt, having written data on it
afterwards, seeing the data retained, it should never later
be lost.

You have a serious problem there, I'd consider it the worst
kind, continually losing data. Only thing remotely close to
worse is having memory errors and thinking you still had the
data, even backing it up but later finding it's corrupt.

Trash the system, clearly this isn't working out. Throw it
ALL away and start out with none of these variables. It
might seem extreme, but honestly I'd do that if the only
choices were that or continually losing an entire drive's
worth of data. That's just hypothetical though, long before
now i'd have reinstalled a slipstreamed Win2kSP4, tried
another IDE cable, another controller, checked the PSU,
removed the removable drive bays, checked the drive
jumpers, replaced the drive. Probably more, I'd never let a
system sit in this state.

I have seen this very thing before when I try to mount two identical
drives, one a clone of the other. In that case the symptoms are
identical - the device shows up but will not mount. That is a logical
device problem, not a hardware problem.

yes that' a logical problem but not necessarily related in
any way.

Hopefully this will be the end of it for a while. I just learned that
Vista is going to be delayed so there went my hopes of rebuilding the
operating system for the foreseeable future.

Bob, Win2k is NOT prone to this problem. You have no
reasonable expectations that any other OS would be a
solution any moreso than a clean reinstall of Win2kSP4. So,
all Vista will do for you is cost a few bucks and slow the
box down some in addition to taking a few Gigs more HDD
space.

I may have set some kind of record in terms of not rebuilding Windows.
I started out with NT4 in 1997 and went thru all the SP and then
loaded Win2K in 2000 as an Upgrade over NT4. I have never actually
installed Win2K from scratch.

Well there you have it... sometimes people install windows
clean to get rid of issues, or at least a preventative
measure. It is incredible that you haven't done this yet.
What a mess the OS must be at this point.

I have done an "in place upgrade" a
couple of times and maybe I should do it again but it is such a bitch
sorting out the things that it screws up.

No, maybe you should just do a normal clean install of
slipstreamed Win2kSP4. It works Bob, and works very well
(I'd rather it than WinXP or Vista).

I keep telling myself that I will not reinstall until Vista because
it's just around the corner. HA!

So in addition to having potential hardware problems you
will introduce a ton of new OS bugs? Never a good idea
trying to install a new OS to cure pre-existing but
unidentified problems.

We can only hope Vista won't allow you to do an in-place
upgrade else we'll be hearing of the same problem for the
next few years. Just fix it, at least do a clean Win2k
install so you have that system shipshape when it comes time
to buy a new box fast enough to run Vista well.

 

[Bob]

Ok but, what if you keep using that other drive in this
situation? Presumably the drive now having trouble, also
seemed to work ok initially?

That's where we differ in our opinion. I do not believe the drive
itself is having trouble. Win2K did not report any problems at the
device level. It was only at the logical level that there was a
problem. I have backup so I am not too bothered if it craps out.

You might be right but it's a premature conclusion to make,
and one that if wrong, could result in a lot of unfruitfull
effort.

Are you saying that Win2K can't even diagnose a device failure
correctly? I am relying on the claim that Device Manager made about
the health of the physical drive. Is it wrong to assume that?

They're not terribly precision devices, these drive bays.
The connectors alone would cost more than the entire
enclosure if that were the case.

I do not have any reason to believe that the removable drive bay is
the culprit.

- Removable drive bay, JUST REMOVE IT!
Get it out of the equation because the only sane way to
tackle your problem is elimination of variables.

I will probably do that if the problem occurs again. But before I do I
will go thru the same ritual I went thru this time in particular using
WD DJG diags, which I believe corrected a corrupted MBR.

As for not removing the removable drive bays entirely, you
are nuts to have not done this already given the problems
you've been having retaining data. What's the point of
being able to remove an empty drive???

I have no reason to believe the removable drive bays are the cause of
any problems.

 

[kony]

That's where we differ in our opinion. I do not believe the drive
itself is having trouble. Win2K did not report any problems at the
device level.

You are putting too much weight on what an OS does or
doesn't "report". At best, an OS simply follows it's
routines and if that fails, the pre-programming return code
that someone who has never seen your drive or system, set to
be returned. It can indicate certain types of problems but
is not a diagnostic in itself to any degree. "Hints" of
what's happening might be a good way to interpret.

It was only at the logical level that there was a
problem. I have backup so I am not too bothered if it craps out.

Sure, logical level. Suppose by some fluke of electrical
nature, the onboard cache on your hard drive is not stable,
is producing errors. What might the result be? The OS sees
an "ok" drive, though it's logically problematic due to the
drive's hardware problem. I'm not implying we should expect
a bad cache chip on your drive, 'tis only one example of a
situation where shades of grey make it inappropriate to make
the sweeping conclusions you are. Perhaps you can take it
as a hypothesis, but now you still have to test that
hypothesis before relying on it.

Are you saying that Win2K can't even diagnose a device failure
correctly?

I'm saying that no operating system on earth can
comprehensively diagnose all device failures, or many kinds
of failures on many kinds of devices. Consider the
complexity of a hard drive relative to, oh a stone or a
transistor or whatever... myriad things can be wrong and the
OS is not meant to diagnose them.

I am relying on the claim that Device Manager made about
the health of the physical drive. Is it wrong to assume that?

Yes

You might weight it as one piece of evidence, but it is not
conclusive.

I do not have any reason to believe that the removable drive bay is
the culprit.

bob, YOUR DRIVE CAN'T EVEN HOLD DATA!

That is clearly a reason to suspect the removable bay.
You don't really care if you lose all your data, apparently.

I suppose that's where we differ, my data is worth more than
the systems it's on. I would care less if the drives were
removable if it added a risk of data loss and it would be
among the first variables to check.

An overview of your situation: You are prematurely drawing
conclusions and not actually isolating any variables. No
wonder you still have the problem.

I will probably do that if the problem occurs again. But before I do I
will go thru the same ritual I went thru this time in particular using
WD DJG diags, which I believe corrected a corrupted MBR.

Your problem is not an MBR problem. If the MBR is actually
corrupt, it is only the result of the problem you are trying
to find. Why continue with the ritual of restoring MBRs
until you determine WHY you are doing it, since you're just
going to perpetually lose data until the actual problem is
corrected?

In short, it is a complete waste of time for you to fix the
MBR. Until you find the problem, the drive is simply
unsuitable for data storage. So is your other drive.

What size are they? Are you sure your system properly
supports 48bit LBA? What service pack level is the OS at
currently? Have you ever ran a 48bit LBA check(er) type
utility to confirm support? Google should find some, one of
which is named (IIRC), "enablebiglba.exe"

I have no reason to believe the removable drive bays are the cause of
any problems.

I have no reason to believe you're ever going to find the
problem if you don't start isolating variables. You do in
fact have a reason to suspect it, and denial will not solve
your problem. It may not be the removable bay but it is
"something" and that something will not change based upon
belief alone.

Why are you struggling so hard against fixing your problem?
Are you in love with your removable bay?

Perhaps a better question is, what DO you believe the
problem is, and if you don't have any specific belief, then
how can you rule anything out? Yes I know you mentioned
"but Disk Management xxxxxxx", which isn't telling you
anything.

Do you understand that if your removable bay has an
intermittent contact, it can still show up ok in Disk
Management so long as it's connected at that moment?

It seems you don't actually care if you solve the problem.

 

[Bob]

You are putting too much weight on what an OS does or
doesn't "report". At best, an OS simply follows it's
routines and if that fails, the pre-programming return code
that someone who has never seen your drive or system, set to
be returned.

Having written many embeddes systems in assembly I would settle for a
return code. But Win2K is completely silent.

Sure, logical level. Suppose by some fluke of electrical
nature, the onboard cache on your hard drive is not stable,
is producing errors.

Then why do I not see those errors happening at other times?

Perhaps you can take it
as a hypothesis, but now you still have to test that
hypothesis before relying on it.

The only working hypothesis based on the data at hand is that the MBR
somehow became corrupted and WD DLG fixed it.

I'm saying that no operating system on earth can
comprehensively diagnose all device failures, or many kinds
of failures on many kinds of devices.

A simple statement from Win2K would go a long way. But Win2K is
totally silent.

Yes

Then why have it at all if it cannot be relied on?

You might weight it as one piece of evidence, but it is not
conclusive.

If I take it into account I have to conclude that the drive as a
physical device is not at fault. The only thing left is the contents
of the drive, eg the MBR

bob, YOUR DRIVE CAN'T EVEN HOLD DATA!

I think that is a bit hysterical. My drive holds data.

That is clearly a reason to suspect the removable bay.
You don't really care if you lose all your data, apparently.

I have backup schemes in place that I use to recover properly.

An overview of your situation: You are prematurely drawing
conclusions and not actually isolating any variables. No
wonder you still have the problem.

When I see positive evidence that the removable drive tray/bay is the
a possible cause, then I will remove it from the equation.

I discovered that because of a Win2K problem that I have dramatically
reduced the incidence of disk corruption. That tells me the problem is
not with the removable bays.

I have installed just about every kind of application and utility
known to computing. I keep clone backups precisely because I have to
recover from a bad install and the easiest way is to go back one
generation, since that generation is only 10 minutes old. However,
many times I am able to uninstall without going back a generation.
What I do find is that the uninstall does not clean out the crap in
the Registry so I have 6 Registry cleaners which I use cautiously to
try to clean things up.

Nevertheless even the cleaners can't get rid of all residue so
Iundoubtedly have something in the Registry that is causing disk
corruption. I found one piece of it the other day and now the disk is
not being corrupted anywhere near as often.

Your problem is not an MBR problem. If the MBR is actually
corrupt, it is only the result of the problem you are trying
to find. Why continue with the ritual of restoring MBRs
until you determine WHY you are doing it, since you're just
going to perpetually lose data until the actual problem is
corrected?

I am trying to find the cause. I think I did actually find something
which was likely part of an application install that the Registry
cleaners can't find because it is not tagged to anything.

In short, it is a complete waste of time for you to fix the
MBR.

You are forgetting that when I did fix it with WD DLG, the disk
mounted properly.

Until you find the problem, the drive is simply
unsuitable for data storage. So is your other drive.

Because I have an extensive backup scheme, I do not worry as much as
you do. I have had corrupt disks a lot more than most people, and I
have always recovered.

I have no reason to believe you're ever going to find the
problem if you don't start isolating variables.

That's what I am doing right now.

But I am working from a different set of assumptions than you are.

 

[Bob]

You are putting too much weight on what an OS does or
doesn't "report".

Here's something I keep getting from Event Viewer when the boot disk
becomes corrupted and I have to mount it as D: so I can run CHKDSK
(which always works to fix the problem):

"The following boot-start or system-start driver(s) failed to load:
PCLEPCI"

WTF does that mean? Even when Windows does speak it makes no sense.

 

[Pen]

Here's something I keep getting from Event Viewer when the boot disk
becomes corrupted and I have to mount it as D: so I can run CHKDSK
(which always works to fix the problem):

"The following boot-start or system-start driver(s) failed to load:
PCLEPCI"

WTF does that mean? Even when Windows does speak it makes no sense.


--

"A politician's neck should always have a noose around it.
It keeps him upright."
-Robert Heinlein

The file you mention is a Pinnacle one having to do with a
PCI card of theirs. However, since you get the same error
which you apparently fix with chkdsk doesn't that mean the
drive is heading out?