1 TB WD Black WD1001FALS suddenly extremely slow

Question

I have and internal hard drive disk, 1TB Western Digital WD Black WD1001FALS, that suddenly got really slow. It just happened as I was using my computer normally for daily tasks, with no sign of problem whatsoever before. The drive is around 5 or 6 years old, I think, and was working great until now.

Everytime I try to access it, the computer just gets really slow, and takes 10 seconds to 30 seconds to list a folder. Sensing that the drive was failing, I've tried a multitude of things:

Running chkdsk /r - Passes the first 3 phases without any problem, gets to 194th sector of the 4th phase and just hangs there for hours and hours.

Boot on Ubuntu Live CD and tried to copy the files to another drive. The speeds are extremely slow, so I just gave up (only 25GB in 12 hours).

While in Ubuntu, run TestDisk, dd_rescue or just dd. All of them ended up with the same result, super slow speeds.

I'm currently using Acronis True Image WD Edition to create a drive image. It's faster than the previous tools, but it's still very slow. I have 750 GB of data that is very important - work related, personal family photos - and I can't lose it, so I really can't give up on the drive yet.

Is there anything I can do to speed up the process? I think the problem is that the sectors aren't marked as bad, so the data can still be read but takes huge amounts of time to do so. I wish I could just skip the problematic sectors automatically, but even using the above tools with settings to skip bad sectors, they keep being read and so the process is exhausting. Should I wait and see if Acronis can finish the backup or try something like a recovery service? I've heard they charge you quite a bit, and I'm afraid I can't pay that expense at the moment if that's true.

Here is a screenshot of the SMART data:

My motherboard is an ASUS P6T Deluxe, running the drives in AHCI mode, using Intel RST 9.6 drivers, everything else is up to date. I'm backing up to a new 3TB WD Green HDD.

Solution 1

You have a warning about Reallocated Sectors Count:

Count of reallocated sectors. When the hard drive finds a read/write/verification error, it marks that sector as "reallocated" and transfers data to a special reserved area (spare area). This process is also known as remapping, and reallocated sectors are called "remaps". The raw value normally represents a count of the bad sectors that have been found and remapped. Thus, the higher the attribute value, the more sectors the drive has had to reallocate. This allows a drive with bad sectors to continue operation; however, a drive which has had any reallocations at all is significantly more likely to fail in the near future. While primarily used as a metric of the life expectancy of the drive, this number also affects performance. As the count of reallocated sectors increases, the read/write speed tends to become worse because the drive head is forced to seek to the reserved area whenever a remap is accessed. If sequential access speed is critical, the remapped sectors can be manually marked as bad blocks in the file system in order to prevent their use.

You should replace the HDD as soon as possible.

Solution 2

Don't replace it! Western Digital have been very naughty in their fast-and-loose determination of which sectors make it to the re-lo table. Clear the damn table and see how things go (AKA do a non-destructive sector scan AFTER this procedure):

Step 1) Make sure the drive can be seen on the bus:

sudo lshw -businfo

Step 2) Get yourself hddsupertool (Unless you follow the manual method) here.

Run as root, select the device, and select the VSC entry.

Step 3) Select this entry:

4) WD royl (Marvel) patch mod 02 (slow fix)

Step 4) You want to first

1) Read the module to a file and create the patch.

followed by

2) Write the patched data back to the disk.

Step 5) Go back to VSC entry, select:

5) WD royl (Marvel) patch mod 32 (slow fix)

and perform entries 1 & 2 again (reading mod, patching, and writing back)

Note: It may tell you it's failed on the final step. It still worked for me.

Step 6) Power cycle the drive

What you've just done: You've cleared the SMART status, and the file relocation list. The re-lo list contains info about which sectors need to be relocated, which in WD's case happens quite nearly every time on a failed read. This means the tiniest vibration, a slight cord stretch, or fairies bangin' on the platters.

Credit: Spildit (Manual Method of doing this) & Scott Dwyer (Author of hddsupertool)

I went from 130KB/s scan i/o speed in testdisk to 11-13MB/s after the fix, and I was able to recover data at +100MB/s from as far back as the very first time I used my similar drive. WD is not the company they used to be.

Key notes: Back your stuff up, c'mon, you should no matter what. If you've done this, then you really have nothing to lose by trying this.

Good Luck!

It really grinds my gears seeing people just agree with the status quo on electronics hardware, especially harddrives, along the lines of "it's bad, replace it." Looking a little deeper, most of my harddrives have not only been rescued thanks to seemingly arcane methods, but most are actually still in service today!

Comments

[Aaron Miller]

The more you beat on it, the likelier it is to fail. Let the backup run to completion, and then see what it was able to retrieve; you don't know until then whether you need to think about spending four figures (USD) on a cleanroom recovery service, or whether you can just replace the drive (for best reliability, with a backed-up RAID 1 array) and go on about your business.

[gronostaj]

... and start making regular backups!

[MRA]

Well, I just got a Blue Screen. It was caused by vidsflt.sys, and seems to be related to Acronis. Not sure what else to do at this point.

[DavidPostill]

@Englishman That's not how Super User works. We don't close questions just because they have an (accepted) answer. Please don't try and do that.

[SmokinGrunts]

Okay guys it wasn't even 2 minutes before the downvote on this. Just tell me why. Sure, the question was from years ago. Someone searching a query that pops this thread up though? Bam, here's a viable alternative (though seemingly unpopular) answer.

[Thalys]

My opinion is "if you can RMA it. you should - else do not use it for anything critical". I have a drive that died... twice so far, and keeps coming back to life for no apparent reason. Where's the documentation for what these commands actually do?

[Pankaj]

@SmokinGrunts: Thank you so much for your answer, now my HDD's copying speed 50 MB/sec from 80 KB/sec, it's super fast :)

[user3210008]

Thank you! My ddrescue's average rate went from 631 kB/s to 165 MB/s after applying this solution! FWIW, I have a WD2003FZEX. I used HDDSuperClone (which now includes HDDSuperTool) with the ATA Passthrough mode, and I only applied the mod 02 fix (after some research it seems that you should only apply the mod 32 fix only if the 02 fix didn't work).