0

My wifes computers hard drive effectively died. All write operation fail and active time is constantly 100%. Drive is currently unmounted as to not slow down the entire system, but still active time shows 100%.

Is there a way for to copy all the data from the unmounted drive to another drive without having to create a backup or mount the drive? Since writing is slow/not working there are many operations that cannot be done.

Or can I mount the drive in such a way that the data related to mounting is stored on another drive?

PC runs on Windows 10.


Update: I got a Ddrescue for Windows (it run through Linux virtual machine). Weird thing happened. Active time is now on 50% not 100%. I have a feeling that there is some application on the drive that will suck up the read/write on it when I try to acces the drive. but I dont know what it might be. My best guess ATM is Epic Games as it was already getting slower and slower to run.

Once Ddrescue complete I will see what happens on the new drive. If same thing happens, then I need to start to remove apps from the drive somehow


Update: Using ddrescue a complete copy of the disk was made on another HDD. Everything works fine now. No data was lost, beside some icon links.

Some things I noticed during the rescue process. Once I opened the computer with the faulty drive connected. Active time was 0%. I immediately opened ddrescue and got to working. During the entire clone/copy process activity never spiked over 65%. No bad sectors were found. My assumption is that the read/write heads were fine and the disks themselves are all good, but the controller of the drive was getting errors either from drivers or physical which causes it to go into some endless loop when trying to directly access it through Windows.

7
  • The drive is dead. I assume you don’t have backups?
    – Ramhound
    Commented Jan 13 at 7:38
  • No backups. It is tehcnically accessible. If i mount it. its just super slow
    – Marko Taht
    Commented Jan 13 at 7:45
  • My suggestion. Get yourself a HDD duplication device. It’s your best chance to get the data of a disk behaving the way you describe
    – Ramhound
    Commented Jan 13 at 7:56
  • Made my sort-of suggestion in the answers, though that's mostly for geeks. I don't know if there is a GUI tool for data recovery, but I think it would be more useful for the general case.
    – mcendu
    Commented Jan 13 at 8:24
  • 1
    A consumer HDD duplication device such as a docking station assumably lacks the abilities of ddrescue and will most likely duplicate the disk in a linear fashion. If the source disk has lots of pending sectors using such a device will stress the source and recover in an inefficient way - wasting time on erroneous zones although lots of readable sectors are available elsewhere.
    – r2d3
    Commented Jan 13 at 9:00

2 Answers 2

2

The Linux world has a tool called ddrescue for recovering data from near-dead drives. You only need to boot into a Linux installation disk to use it – you don't need to install Linux from there.

The SystemRescue distribution comes with ddrescue on its ISOs. Burn the ISOs on a CD or a USB thumb drive, then reboot into the CD/USB thumb drive to gain access to ddrescue.

ddrescue's basic usage is as follows. Ensure /dev/sdx is the bad disk with all your precious data, and that /dev/sdy is a good disk larger than /dev/sdx (Thanks Kamil Maciorowski):

# ddrescue /dev/sdx /dev/sdy data.map

You may need to fix /dev/sdy's partition table and run checks (fsck or CHKDSK) on each partition after recovering.

Note that its usage requires some Linux knowledge and command-line basics. There is no drive letter like in Windows, and you have to name your hard drive as something like /dev/sda or /dev/nvme0. If you are unsure which is the bad drive, you can use lsblk to get a clue.

5
  • 2
    And then what? Will Windows 10 be able to do anything useful with data.img? How to proceed? For average Joe a better choice may be to rescue directly to a block device, because there's a chance the target device will "just work" in Windows. There are pitfalls and edge cases even then. Commented Jan 13 at 8:47
  • @KamilMaciorowski My own workflow, sort of. Edited.
    – mcendu
    Commented Jan 13 at 9:08
  • I vaguely recall that windows has some image mounting abilities since 10, else the tools you need to mount it are in linux
    – Journeyman Geek
    Commented Jan 13 at 9:30
  • @JourneymanGeek At least ISO9660 as far as I can recall. Probably also FAT. Likely not MBR or GPT images which can contain multiple partitions.
    – mcendu
    Commented Jan 13 at 9:42
  • And then what: Then you can try mount disk image but often you'd need to use a file recovery tool that can work with a sector by sector disk image, examples UFS Explorer, DMDE .. Commented Jan 13 at 11:38
1

assuming this is a rotating disk drive, likely the more the drive is used, the more it will get damaged. (there is an avalanche failure mode due to... reasons.) so do not power it while it is not being rescued.

you can try rescuing it: you need a place to store the rescued contents (space in a disk for a file as big as the failing disk. you could also use a new disk at least as big as the old one, but you still need another disk to store some extra files.)

  1. first download and flash a live linux distro image to a pendrive. i recommend linux mint cinnamon.
  2. check that you can boot linux mint.
  3. power off, connect the failing drive and rescue destination drive(s) to the PC, then reboot linux mint.
  4. use the "Disks" application to check that all drives are visible to the PC.
  5. i assume you do not have linux swap partitions in the failing disk (you were not using linux before). but if you do, use "Disks" to unmount/stop the swap partitions there.
  6. open a terminal and type ddrescue to check if ddrescue is installed. if not, connect to the internet then type sudo apt install gddrescue to install it, then retry.
  7. follow some ddrescue how-to, such as this one: https://linuxconfig.org/how-to-repair-and-clone-disk-with-ddrescue

notes:

  • ddrescue produces two outputs: a backup of the readable parts of the disk, and a log specifying which areas could be read and which could not.
  • it may take almost forever to backup a failing drive: the more damage, the more time.
  • when the backup is done, you will need to reconstruct the readable parts of the disk into useful data. (some tools help: ddrescueview, ddrutility, etc. you can install them with sudo apt install ....)
  • but forget about magic: most data could be lost already.

good luck!

1
  • recovered all data :)
    – Marko Taht
    Commented Jan 13 at 13:59

You must log in to answer this question.

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