flaky WD Black 2TB drive

[BLinux]

I have a 2TB WD Black drive that I got as an RMA replacement. After I got it, I forgot to test it out and just put it on a shelf of spare hard drives. Recently, when I needed it, I discovered it is a bit flaky.

At first, the drive seems to work fine. I formatted it with XFS and started copying data to it. After several hundred GBs, it started to write really slowly. And I noticed in the SMART data, that current_pending_sectors was incrementing, eventually up to 80+ from 0. I stopped the data transfer, and decided to just wipe the drive with zeros using dd like:

dd if=/dev/zero of=/dev/sdX bs=100M

After doing this wipe, the current_pending_sectors count goes back to 0 and everything seems fine. I again format the drive and I'm able to use it for a little while, but then pretty much the same thing repeats itself.

Not sure what would cause a hard drive to behave this way? And is there some remedy for it? When I got it as an RMA from WD, it was basically a new drive. By now, the warranty has expired, so kind of a bummer that a new 2TB drive is useless. suggestions?

 

[keybored]

WD has some utilities on their site that can do diagnostics, zero out the drive, etc. Have you tried running any of them to see if they throw up any red flags?

I had a flaky WD Green that would just freeze when I tried to zero it out using WD utilities and this would always happen towards the very end of the process. SMART data showed 10 reallocated sectors. I partitioned it such that the last few GB was left unused and it's been running fine ever since with no freezes no matter how much data I sent to it. And no new reallocated sectors showed up since that time either.

 

[andrewbedia]

you've unfortunately got a doorstop now. current pending sectors should never be non-zero.

I hear 2TB Hitachi UltraStar drives are ~$35 or so on ebay.

 

[BLinux]

WD has some utilities on their site that can do diagnostics, zero out the drive, etc. Have you tried running any of them to see if they throw up any red flags?

no, i don't have a Windows machine handy. does their utility do more than read SMART data? can it attempt to repair the drive? just wondering if it is worth the trouble of installing Windows to try it...

 

[keybored]

no, i don't have a Windows machine handy. does their utility do more than read SMART data? can it attempt to repair the drive? just wondering if it is worth the trouble of installing Windows to try it...

It's been a while, so I don't recall all the options. But I don't think you need Windows. There should be a DOS based version that you can run off of a USB stick with Free DOS or another free alternative.

 

[BLinux]

you've unfortunately got a doorstop now. current pending sectors should never be non-zero.

I hear 2TB Hitachi UltraStar drives are ~$35 or so on ebay.

yeah, not an issue to get another drive - just bummed that a perfectly new drive sat on the shelf and has become useless.

as i understand it, "current_pending_sectors" can increment and eventually become reallocated sectors, no? that's just part of the normal operation of the hard drive as it accumulates defects over time and works around it.... in my case, i get a lot of current_pending_sectors that then go away with a wipe, but never get the actual reallocated sector.

 

[BLinux]

It's been a while, so I don't recall all the options. But I don't think you need Windows. There should be a DOS based version that you can run off of a USB stick with Free DOS or another free alternative.

I have this drive connected to a USB3-SATA dock to my laptop. Do you know if FreeDOS can work with that? If so, I might give that a try... I think I may have a FreeDOS USB stick somewhere that I use to update mobo bios.

 

[keybored]

I have this drive connected to a USB3-SATA dock to my laptop. Do you know if FreeDOS can work with that? If so, I might give that a try... I think I may have a FreeDOS USB stick somewhere that I use to update mobo bios.

I doubt WD utilities will work through a USB adapter. I vaguely recall that you have to have a SATA connection. Not only that but you have to switch your SATA controller to IDE mode in BIOS before you run the utility. Although I might be mixing that up with "wdidle". There should be a readme file with all the requirements.
Even if WD didn't have such a requirement, I would try to run the drive directly off of a SATA connection just to eliminate any possibility of the USB adapter interfering in some way and causing the issues you're seeing.

 

[BLinux]

I doubt WD utilities will work through a USB adapter. I vaguely recall that you have to have a SATA connection. Not only that but you have to switch your SATA controller to IDE mode in BIOS before you run the utility. Although I might be mixing that up with "wdidle". There should be a readme file with all the requirements.
Even if WD didn't have such a requirement, I would try to run the drive directly off of a SATA connection just to eliminate any possibility of the USB adapter interfering in some way and causing the issues you're seeing.

ok. thanks. i guess i'll pass on the WD tool then. the usb3-sata dock is not causing the issue, all my other WD drives, including Blacks and RE drives, all worked fine except this one in this dock.

 

[andrewbedia]

yeah, not an issue to get another drive - just bummed that a perfectly new drive sat on the shelf and has become useless.

as i understand it, "current_pending_sectors" can increment and eventually become reallocated sectors, no? that's just part of the normal operation of the hard drive as it accumulates defects over time and works around it.... in my case, i get a lot of current_pending_sectors that then go away with a wipe, but never get the actual reallocated sector.

Any time reallocated sectors happen --> send it back for warranty. That shit doesn't happen to a healthy drive. The exception I'd say would be less than 10. I've seen it once or twice where a hard drive gets ~2-3 dead sectors early on and then lives an otherwise normal life span.

 

[BLinux]

Any time reallocated sectors happen --> send it back for warranty. That shit doesn't happen to a healthy drive. The exception I'd say would be less than 10. I've seen it once or twice where a hard drive gets ~2-3 dead sectors early on and then lives an otherwise normal life span.

actually, i've seen it a lot, where a hdd gets a few bad sectors and then never gets anymore. but my experience goes back to the days of traditional SCSI stuff... i don't know how it applies to SAS/SATA stuff... but I have seen a few SATA drives that did that. but yes, if a drive exhibits that kind of stuff and it is under warranty, I will get it replaced.

in any case, this drive is no longer under warranty.. like i said, it was a RMA replacement that got put on a shelf - yeah, i should have tested it when i got it but I was probably too busy at the time.