"My Passport" returning "Device Fault" status

[Firubat]

Trying to run SpinRite on a WD My Passport external drive, and I get the following error message:

This drive has just taken itself offline

The drive is now returning "Device Fault" status. It must be "power cycled" (shutdown and restarted) to clear this condition & perhaps resume operation. Device Fault occurs when a drive encounters an exceptional condition from which it cannot recover. This could be transient or permanent and it might only occur when SpinRite is working on a specific sector or region of the drive. It may be possible to resume SpinRite past this sector or region. Unfortunately, Spinkite cannot do this on its own since once this occurs power cycling is required.

I tried rebooting and try again, but same error, also when jumping forward to different locations the error repeated sooner or later. I also tried the spinrite forcebios option I saw someone suggested, same result.

Is there something else I can try?

 

[DanR]

Trying to run SpinRite on a WD My Passport external drive

That would be a USB drive. Correct? Thus SpinRite would be using BIOS access as it's only option. Thus the ForceBIOS switch would have changed nothing for you.

You did not specify which level you were trying. Presumably Level 2 or Level 3?

I would suggest trying a Level 1 run. Level 1 is just one simple read, no re-tries and no writes. If a sector does not read it is marked with a U in the Graphic Status Display screen.

BIOS access uses a transfer rate of 256 - already quite benign. If a normal Level 1 run takes the drive offline then you might also consider trying the XFER switch with values of 128, 64, 32, 26, 8

spinrite level 1 xfer 128
Or
spinrite level 1 xfer 64
Etc.

To see if something works.

 

[Firubat]

Yes it's a usb drive. I was trying level 2.
I tried now to start at 50% and it managed to almost reach the end, but even going back to 40% gives the error.
I tried now using the level 1 xfer 8 option but it didn't manage to go through.

BTW I now notice that despite the message, if I just exit spinrite and enter again, without rebooting, I am able to restart the operation.

 

[DanR]

Hate to sat it . . . but . . . methinks that drive is toast.

 

[PHolder]

Not that this will help you @Firubat , but relating an experience I had with SpinRite earlier in the development phase of 6.1. I had a drive that would crash and take itself offline when stressed, but it managed to get through the section when the XFER was low enough. (I'm talking like 32, so you've already tried 8 and not succeeded, so it's somehow worse than my drive was.) In my case, my drive was connected via SATA, and I did require to power cycle the machine for the drive to start responding again.

So, if it were possible to shuck the drive and if it has a SATA inside, you MIGHT get a different experience, and you might not. At this point, you may have little to lose by trying if you can.

 

[Firubat]

I wonder if I can just partition it and use the good part, as it seems that about 50% of the drive is fine, and that's 500GB worth of storage. I think it's a spinning drive, so there might just be physical damage on parts of it.

 

[PHolder]

I wonder if I can just partition it

Well, of course you could try and see how it goes, but I wouldn't put much faith in the durability of a drive that is already acting flaky. So far as I know, all the drives manufactured today have some sort of embedded controller that is basically a HDD OS. It runs and makes its own decisions on how to optimize its operation in support of giving you data durability that you expect. Unfortunately, the inner workings are highly proprietary, and all you get to interact with the drive is the SATA commands, or in your case, even less, a USB to SATA command adaptation layer. I think this is where things are probably going wrong for you. When the drive errors out, the USB translation of the SATA error is basically "oops, I'll stop now". And it possibly reboots the USB adaptation layer, hoping for better luck next try. Unfortunately SpinRite only really speaks BIOS, SATA (and IDE) and so is relying on FreeDOS and your BIOS to do the best it can with making the USB device stay in line with its expectations. If the drive is misbehaving, and the USB layer misrepresents what that means to the BIOS, there's not a lot of hope for SpinRite to make sense of the results.

 

[Firubat]

Well, of course you could try and see how it goes, but I wouldn't put much faith in the durability of a drive that is already acting flaky. So far as I know, all the drives manufactured today have some sort of embedded controller that is basically a HDD OS. It runs and makes its own decisions on how to optimize its operation in support of giving you data durability that you expect. Unfortunately, the inner workings are highly proprietary, and all you get to interact with the drive is the SATA commands, or in your case, even less, a USB to SATA command adaptation layer. I think this is where things are probably going wrong for you. When the drive errors out, the USB translation of the SATA error is basically "oops, I'll stop now". And it possibly reboots the USB adaptation layer, hoping for better luck next try. Unfortunately SpinRite only really speaks BIOS, SATA (and IDE) and so is relying on FreeDOS and your BIOS to do the best it can with making the USB device stay in line with its expectations. If the drive is misbehaving, and the USB layer misrepresents what that means to the BIOS, there's not a lot of hope for SpinRite to make sense of the results.

Thanks for the detailed answer! So I'll see if I can use it for something non-critical, or maybe see if there is a direct SATA connection I can plug into.
Cheers!

 

[peterblaise]

"... SpinRite on a WD My Passport external drive, and I get the ...
error message: This drive has just taken itself offline ... The drive
is now returning "Device Fault" status. It must be "power cycled"
( shutdown and restarted ) ..."​

Manually:

SPINRITE NORAMTEST SKIPVERIFY LEVEL 2 DYNASTAT 0 NOREWRITE

... or in a batch file:

SPINRITE NORAMTEST SKIPVERIFY LEVEL 2 DYNASTAT 0 NOREWRITE AUTO EXIT
READSPEED
... repeat.

Worked for me.

After 2 weeks of constantly re-running that command in a batch file, it
went from 14 hours end-to-end to "only" 7 hours end-to-end, apparently
repairing a lot, enough for ddrescue to get all the data I needed.

Otherwise, with DynaStat on or Rewrite, the drive would go offline,
requiring a power cycle and reloading of SpinRite.

 

[PHolder]

DYNASTAT 0

Every time you give this advice @peterblaise you NEED to remind people it will cause data loss.

 

[peterblaise]

Peter Blaise wrote:

Manually:​

​

SPINRITE NORAMTEST SKIPVERIFY LEVEL 2 DYNASTAT 0 NOREWRITE​

DynaStat default or DYNASTAT 1-99 throws the drive off line.

DYNASTAT 0 keeps the drive online for the entire scan, end-to-end.

N0REWRITE prevents data loss.

For me, a significant recovery in 2 weeks, a 14-hour scan dropping
to 7 hours.

No data loss - cool or what?

Nothing but recovery, recovery, recovery.

SpinRite is terrific that way.

All I can do is report my experience.

I hope this helps others know the road ahead by reading a report from
someone coming back.