Slow hard drive with possible bad sectors

Question

My computer recently started to stutter a bit lately and one of my Seagate hard drive became really slow when opening folders, running programs from the drive etc., ultimately ending with the computer restarting. Upon reboot it went right into the Windows 10 chkdisk on boot. It was very slow as well, so I ended up leaving it over the night. Before I left the computer I saw there was a bad sector mentioned. Next day I was a bit more relieved when I saw the logon screen, and I found the hard drive in Windows explorer. First thing I did was to start copying everything, but the transfer speed is very slow. It jumps to between 1 and 3 mbps, until it drops down to 0 again. In one day I was able to copy around 20 gb.

With the hard drive being this slow I guess it's a sign of the drive slowly failing. I checked the computer management, and it seems the disk has changed from Basic to Dynamic for some reason I don't know after running chkdsk.

I would appreciate some advice on what's the best idea right now. My first thought is to leave the computer on and just keep copying files. Or is it better to take breaks every now and then? And with the disk being 1.8 TB and with this speed it will probably take a month until everything is copied, and I'm a bit afraid the drive will fail before it's completed. And will the heat produced from the HDD possibly have a bad impact on its already failing condition? Or is there better ways to take backup, or are there other possible causes for the hard drive being slow? I have tried swapping cables and hard drives to check.

Answer 1

First, use any of the free disk imaging utilities, such as Macrium Reflect Free to copy the whole disk. If possible, boot from USB rather than the HDD, to avoid writing to the disk! Reflect, for example, can make a boot USB drive using WinRE (AKA WinPE). This has some large benefits:

• Imaging applications use direct disk I/O, rather than copy a file at a time. On a healthy HDD, I find imaging 2 TB of files takes about 3 or 4 hours, and about the same additional time to verify the image (Important! A bad image is not very useful).
• Since the entire disk is saved, you can restore that image to a drive of the same size or larger, and have your system working as before.
• If you're booting from USB, there's no need to write to the bad HDD, possibly causing more damage.

Caveats:

• If the disk has many bad sectors, there will be many retries to read each such sector, slowing imaging.
• If a complete image cannot be made and verified, try to mount the partial image and view files; it might still salvage most of the disk.

If there are files you cannot salvage, that you must have , get an estimate from a commercial data recovery service, but expect it to be costly.

In the future, make periodic images of your HDD to avoid trying to salvage it after there's an issue.

Answer 2

Because of the slow speed and size of drive, you may wish to take it to a disk repair agency to see if they can copy the files, or help you in any other way.

Just leaving it to copy in the state it is in will (very likely) accelerate the damage and then subsequent failure.

Answer 3

The first thing to in such a situation is to get SMART parameters. When you do this under linux using smartmontools that does not stress the drive too much.

You most likely have broken sectors. As the source of dammage is unknown the best solution is to use a fault-tolerant copy tool on sector level.

Copying files on file system level is stressful. Using the Macrium software as a duplicating tool is not a good idea because it has not been designed to handle bad sectors.

ddrescue (Linux-based) is the specialist.

Running chkdsk is a bad idea as well. This is something you can do with your second duplicate after copying the faulty source to a target and copying the physically healthy target to another drive for test and play purposes like using chkdsk.