Serious device fault- my first experience with this dialog box

[markav2]

I attempted to do a full ghost image of a spinning laptop drive used in an industrial control. Ghost reports that there are damaged sectors and it will not complete an image. I have release candidate 5 and I decided to let it loose to see if it could recover the bad sectors. I could not even get that far because the below dialog box came up. After several attempts at restarting it, I received the same dialog box every time. The weird thing is that I can easily recognize the drive in Windows and even copy data from the drive. Other drives that I have had that experienced failure were hard to read and sometimes not recognized.

What is triggering this dialog box?

Mark V

 

[PHolder]

What is triggering this dialog box?

So I'm gathering you're refusing to either read the dialog, or to believe what it says. Drive firmware is a black box... the drive makes the decisions it makes for the reasons only it knows. All SpinRite can do is report reality... and the reality is that drive is unhappy with its lot in life and has decided to stop being a drive for the time being. You can power cycle and hope it resumes being a drive, knowing that if you don't have a backup of the data on it, that should be your focus. SpinRite can be restarted beyond the problem area, and hopefully the rest of the drive will behave better.

 

[DanR]

What is triggering this dialog box?

@markav2,

The “message” in the red pop up box would seem to be self explanatory. This is a very troubled drive that is shutting itself down to protect both itself and the data on the drive. This is a situation that SpinRite 6.1 RC5 currently cannot cope with.

This may leave you with some options, however. If you can see the percentage at which this fault occurs, then as PHolder suggested, you could start SpinRite at a point beyond the trouble spot to see if SpinRite can continue on from there. If SpinRite encounters a bad unrecoverable sector such that a drive shut-down is not triggered, SpinRite would issue a Write to force a re-allocation of that bad sector (which may or may not subsequently trigger a drive shutdown)

CAUTION: This could result in a permanent loss of any data in that bad sector. It should be attempted only if you are not concerned with data loss.

CAUTION: This may NOT be practical given the apparent troubled nature of this drive.

Or, you could wait for SpinRite 6.1 RC6, which we may or may not see by Dec 31, 2023. This SpinRite release may give you the option of leaving a bad unrecoverable sector as is and proceeding to the next sector. This approach could recover any recoverable sectors and possibly allow more recovered data to be copied off of the drive afterwards. RC6 may also have to ability to not try so hard to recover a bad sector, thus very possibly avoiding a drive shutdown.

This drive would then be suitable only for testing by future versions of SpinRite to see if/how they might cope with a drive such as this.

 

[markav2]

@markav2,

The “message” in the red pop up box would seem to be self explanatory. This is a very troubled drive that is shutting itself down to protect both itself and the data on the drive. This is a situation that SpinRite 6.1 RC5 currently cannot cope with.

This may leave you with some options, however. If you can see the percentage at which this fault occurs, then as PHolder suggested, you could start SpinRite at a point beyond the trouble spot to see if SpinRite can continue on from there. If SpinRite encounters a bad unrecoverable sector such that a drive shut-down is not triggered, SpinRite would issue a Write to force a re-allocation of that bad sector (which may or may not subsequently trigger a drive shutdown)

CAUTION: This could result in a permanent loss of any data in that bad sector. It should be attempted only if you are not concerned with data loss.

CAUTION: This may NOT be practical given the apparent troubled nature of this drive.

Or, you could wait for SpinRite 6.1 RC6, which we may or may not see by Dec 31, 2023. This SpinRite release may give you the option of leaving a bad unrecoverable sector as is and proceeding to the next sector. This approach could recover any recoverable sectors and possibly allow more recovered data to be copied off of the drive afterwards. RC6 may also have to ability to not try so hard to recover a bad sector, thus very possibly avoiding a drive shutdown.

This drive would then be suitable only for testing by future versions of SpinRite to see if/how they might cope with a drive such as this.

I do not even get a chance to scan the drive at all. The message pops up as soon as I select the drive. As I said above though, I can access the drive from within another copy of Windows and the drive seems to act normally, allowing me to copy files off of it.

 

[PHolder]

I do not even get a chance to scan the drive at all.

SpinRite exercises the drive during the detection phase. I think it tries to read special places like the first and last sectors, and I think it actually tries a write toward the last sector as well. @Steve will no doubt chime in with more specific details. In any case, something unfortunate is up with your drive that a normal OS boot doesn't detect, but SpinRite does detect (well evoke, really) during drive discovery. As I said before, make sure your backups are good. There are some S.M.A.R.T. utilities, as well as utilities from the drive manufacturer that can potentially give you some statistical insight into what is upsetting your drive.

 

[DanR]

Ah ... I see now that the drive apparently makes it past the drive detection screen, where drive exercising does occur, and takes you to the "Select Drive(s) For Operation" screen. And then the red popup occurs before proceeding past this screen? When you select the drive? Ore before you can do anything on this screen?

 

[Steve]

Hi @markav2:

Hmm.

During its initial drive recognition process, SpinRite verifies its ability to read and write to the drive by doing so at the very end of the drive. It also determines the drive's mapping to the BIOS using several strategies, including reading and hashing the first sector of the drive and comparing that to hashes created through the BIOS. And it verifies its ability to transfer entire 16MB blocks of data in the middle of the drive while timing that to obtain its estimate of the time required to read scan the drive. And since every or any one of those actions might be thwarted by a troubled drive, they are all extremely fault tolerant with the ability to try at many various locations on the drive.

Was the drive's line-item in that initial listing shown in RED? And did you then attempt to select that RED line-item in the drive selection dialog? That would seem to be the most probable sequence of events. If so, that would mean that at some point during SpinRite's confidence checking it's touching some sector that literally "crashes" the drive. The bad news is that this behavior is more common on spinners than we would like. When a drive declares "fault" it will not respond to resets and it truly needs to be powered down and back up.

After several attempts at restarting it, I received the same dialog box every time

Just verifying that you did power down each time? A true full power cycle is needed to clear that "Device Fault" status.

I would very much like to find a way to induce SpinRite to at least make an effort with that (and other similar) drives. The only way I can see to do that would be to offer a form of bypass that would authorize SpinRite to skip all of that initial work, some part of which is knocking that troubled drive offline. One thought would be that the Device Fault dialog could display a short token (obtained from a hash of the make/model and serial number of the drive) that when added to the command line would authorize SpinRite's use on that drive without any of the initial testing that caused the drive trouble.

It would still be the case that when SpinRite does touch whatever that sore spot is, the drive would likely crap out... but it might also be that before it got to that point, some useful data recovery could be performed. And also, the point where the "crap out" occurs could be noted and SpinRite could then be restarted, again using that token, past that "don't touch me there" location in order to continue working on the drive.

So if, for example, the "sore spot" is at the very end of the drive, SpinRite might be able to help everything else, only crashing out when it reaches the end.

 

[Mervyn Haynes]

It would still be the case that when SpinRite does touch whatever that sore spot is, the drive would likely crap out... but it might also be that before it got to that point, some useful data recovery could be performed. And also, the point where the "crap out" occurs could be noted and SpinRite could then be restarted, again using that token, past that "don't touch me there" location in order to continue working on the drive.

"Sore spot", "Crap out", "Don't touch me there"? Are these technical terms? :-} Or is a bit of Peter rubbing off?