Menu

Trying to recover a failing drive that was encrypted with Veracrpt

James
2026-02-04
2026-03-24
  • James

    James - 2026-02-04

    I have an internal 4TB HDD encrypted with veracrpyt that I use as an extra storage drive. One day a database program I had running on it started to freeze while it was still writing to the drive along with anything related to the drive. I panicked because I couldn't even open my task manager in Windows and did a force shutdown on my PC and when I restarted, I heard a couple of clicking noises before going back to normal. When I got back into Windows I noticed Windows doesn't recognize the drive anymore and when I check disk manager, it gives me the option to initialize the drive. I didn't because I wasn't sure I'd mess anything up.

    I then tried mounting the drive through veracrypt and when I go to select my drive I noticed the drive still shows up but its missing header information or whatever that usually goes along with drive selecting option. But when I try to mount it veracrypt gives me this warning that the drive can't mount due to is failing and ergs me to back it up. I checked crystal disk program and it showed that I had a yellow warning with 30 something bad sector counts.

    So I eventually learned about cloning failed drives bit by bit since it should work even for encrypted drives however I kept running into issues with the programs needed to do this.

    One program I tried was a updated fork of HDDSuperClone called OpenSuperClone which was suppose to be a continued up to date version of HDDSuperClone since the owner of that program left. I ran a Live USB version of that program. Its Linux based and simple enough to run however whenever I try to clone the drive I kept getting errors about timings.

    I eventually gave up and went to try another Linux base program called DDrescue and the setup was fine however once I ran it I was seeing a lot of "read errors" increasing by the second and "error rate" fluctuating at around 650-700 kB/s with cloning/rescued percentage staying at 0% after 10 or so minutes. I was using google's AI for help and mentioned this wasn't a good sign.

    So I could only chalked this up as a hardware failure that can't be fix through software methods but I never had any 2nd hand human opinion and am wondering if there's anything else I could try to save this drive before I chuck it? Both Linux and Windows still somewhat recognize the drive. Can check the status of the drive and even list the same failed 30 something sectors, its just whenever I start any cloning method, I'd get some kind of issue.

     
  • minesheep

    minesheep - 2026-02-05

    While decryption may still be possible (using the backup heeader) You should be careful if that drive contains important data. Cloning is a good idea, but if a drive is unable to provide any useful data, recovery via software may be impossible. If data on that drive is worth spending in data recovery services. you can ask they for trying to get image, You can then decrypt that image yourself if it has at least one intact header (main one or backup one). However data recovery is often expensive, having a backup (on another drive) is cheaper recovery method, does not help here, but maybe in the future.
    Ps that drive may have stuck motor or head if it can't provide any data.

     
  • ledgeri

    ledgeri - 2026-03-24

    I know i am almost 2 months late, but in the case the disk is still in your hands:
    What i would do a disk what has similar problems, and i do not care about the data, to get a program, like the HardDiskSentinel, or something similar, what overwrites a sector and reads back the result for "readability quality", let it ran through (at least one overwrite, to hide the previous life). + can force HDD-internal deep-scans
    Sometimes an "OS-less" check on a drive can force it, to self heal, meaning it starts using it's preserved area, 30-40 secrors maybe can fit into it. So the drive can kind of self heal...
    If it can not fix itslef, but the problem is "visibly on only one part of the drive", then i still would use the drive, but "partitioning around" the problematic area (widely enough).
    Both would make a perfect dump drive for stuff what good to have around withouth needing to download it again, but still not a problem to if all goes in one click.
    Sure that kind of a disk always can be worse, and can cause OS instabilities, so keep that in mind. Also of course all above mentioned normal use "thought with encryption in mind" :).

    If the fault is already spread, or spreads (HDS is a good monitor tool for that, because monitors everithing what is S.M.A.R.T. data), still a good weeked project to take it apart, and look at the shiny bits :)

     

    Last edit: ledgeri 2026-03-24

Log in to post a comment.

MongoDB Logo MongoDB