0

I've experienced the click of death on my WD3200JS and I'm trying to recover some data from it. The situation is as follows:

  1. I hear the click od death during normal PC operation. The PC freezes (it's my system drive).
  2. I restart the PC, the disk spins up and starts clicking.
  3. I google the problem (haven't heard of the click of death before).
  4. Knowing I can't afford data recovery, I try a quick freeze, take it out, wait 5 minutes. The drive does indeed start up but gets back to clicking after about 2 minutes.
  5. This time I put it in the freezer for about two hours. After taking it out I don't wait and plug it into my PC - this time the drive fails to spin up properly. I figure I should have waited after all, so I wait a bit and plug it in - click click.
  6. I give up.
  7. On the next day, feeling lucky, I plug the disk in to a running PC (without any freezing, I just took it off the table) - the disk works for a few minutes, allowing me to copy some crucial data. WTH??

And now my question is - is there a slight chance this is being caused by a PCB malfunction, in which case I should try replacing the PCB? Or should I just keep trying to freeze it until the disk gives up? The data on the platter appears intact, everything I copied was in perfect shape, but it was very little and I'm afraid of running this disk if there's a risk of damaging the data. Who knows, maybe in a few years I'll be able to afford data recovery from it...

Improve this question
14
  • 1
    What does the disks SMART data say about its current health status?
    – Frank Thomas
    Commented Sep 30, 2015 at 15:02
  • Hard to say, the disk isn't really recognized (by Windows or BIOS) while it's clicking, and on the two occasions it WASN'T clicking, I was frantically copying data off of it. Do you think I should try to check the SMART data instead of grabbing the files while I have the chance?
    – Mike Mayday
    Commented Sep 30, 2015 at 15:14
  • 1
    If you can slow the rate of temperature rise, it might keep running longer, so mount it in an external cradle and locate it in a cold air-stream (ideally, air conditioning output). Until you can afford data recovery, learn the lesson that you cannot afford not to back up regularly.
    – AFH
    Commented Sep 30, 2015 at 15:17
  • Consider using the Linux utility ddrescue for a resumable bit-by-bit copy to another drive.
    – Steven
    Commented Sep 30, 2015 at 15:20
  • 1
    I don't understand the question. How will replacing the PCB solve a mechanical problem? If data recovery is a high priority you are better of hiring a data recover company. If you perform one of those "home solution" your chances of success with professional services will decrease by a significant amount in my opinion.
    – Ramhound
    Commented Sep 30, 2015 at 15:22

3 Answers 3

Reset to default
1

I'm sure you already know this, but the disk is dying.

Based on your SMART stats, in particular the Uncorrectable sector count and Current pending sector count your platter surfaces have been damaged, but the drive appears to be mechanically sound. This means that you stand a good chance of recovering the non-damaged data. There are never any guarantees with data recovery, but it could be worse.

I recommend you use ddrescue (following the approach I suggest here) to image the disk to a new drive. ddrescue will attempt to recover the whole image (both damaged and undamaged sectors/blocks). Note that ddrescue can take days to complete on large, heavily damaged disks.

Once you have an image on solid media, you can attempt to mount it. if mount works, and the filesystem is whole, you should be able to access most of the file data, and do with it what you will. If the filesystem doesn't mount however, or appears to be empty, you will have to fall back to file-carving utilities like Photorec, recuva, remo, Easus, etc. These tools scan a hard disk (in this case, the image you created) in raw mode, and find patterns that indicate the beginning an ends of known filetypes, so they can be recovered even though the filesystem data has been lost.

Just remember, never recover files to the volume you are recovering them from, as it will inevitably overwrite other data that you will wish to recover.

Improve this answer
0

The freezing method is controversial, but I have successfully used it on an old 40GB so in the right circumstances (like yours), its worth recommending.

Problem: Freezing the drive is fine until you take it out and spin it up, it rapidly warms up and can cause condensation internally. Water droplets running around on the platter or ICs is NOT what you want.

Solution: Do the recovery entirely INSIDE the freezer! Laptop on top of the freezer, USB drive interface and 12V power rail fed into the freezer with the door closed over it. Leave the power off so the drive is cabled but powered off for 2-3hrs, then boot it and pull data as fast as you can. Obviously have a target drive ready to store what you can read. If you're lucky you'll get 45 minutes of read time but it may be a lot less.

Improve this answer
0

Alright, so here's an interesting thing: I've plugged the disk in again for the hell of it (without any freezing)... and it's working again. I've managed to copy all the data with no problems. It appears I won't be trying ddrescue, though to all people with a similar problem, I'd suggest it rather than freezing.

Many thanks for your answers, if I ever have a situation like this again, I'll panic a lot less ;)

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .