Question Remap issues with WD Blue HDD ?

[overclocler14]

Hi everyone, over the last few weeks I've been pulling my hair out and decided I must be missing something. Prepare for a long read since this situation is insanely difficult to explain.

Basically I got a 1 TB WD Blue 2,5 inch HDD from my friend to take a look at. The exact model is WDC WD10JPCX-24UE4T0 - nothing fancy, sadly it's only 5400 RPM.

She wants to expand her game storage but with the prices od SSDs currently it's a no-no. The drive was originally in a Lenovo laptop from around 2017 and it was replaced with an SSD after two years because it became unbearably slow. Couple of months ago she bought a gaming laptop but it has only 512 gigs of nvme so she keeps games and other stuff on an aging, external 1 TB USB drive - not an optimal solution and it is not possible to install this drive as an internal drive, since it has the USB controller integrated - no SATA connections.

She remembered she still has the old drive and asked me if I could "do my computer magic" and make it usable, so that she could install it in her laptop. So, expecting only heavy fragmentation, from what she told me, I took the drive and started my standard diagnostic procedure.

Checking the drive's S.M.A.R.T. i noticed that there were 6 pending sectors, strange, since the drive has only around 3500 hours of use and spent more time in a drawer than working. A quick check revealed over 30+ % fragmentation of the system partition - no wonder it was insanely slow. But it was going to be formatted anyway so I did not care.

I then proceeded to do a full surface scan using the Victoria software - have been using it for many years and it's been great at diagnosing drives, it works with basically everything. The scan revealed that the drive is mostly healthy, the only problematic area being around 21 GB mark (LBA 41 800 000), where 8 UNC (CRC) appeared. I tried remapping and all 8 sectors seemed to be successfuly remapped according to Victoria. The pending sector count decreased to 0, however, strangely, the realocation count did not increase (but I saw this happen in other drives i successfully repaired). This is the log output from Victoria:

Spoiler

19:59:09 : Recallibration...
19:59:10 : Recallibration... OK
19:59:11 : Starting Reading, LBA=41 000 000..1 953 525 167, FULL, sequential access w. REMAP, tio 10 000ms
19:59:23 : LBA 41 849 448 try REMAP...
19:59:27 : LBA 41 849 448 try REMAP... complete
19:59:28 : LBA 41 849 449 try REMAP...
19:59:29 : LBA 41 849 449 try REMAP... complete
19:59:30 : LBA 41 849 450 try REMAP...
19:59:30 : LBA 41 849 450 try REMAP... complete
19:59:31 : LBA 41 849 451 try REMAP...
19:59:32 : LBA 41 849 451 try REMAP... complete
19:59:33 : LBA 41 849 452 try REMAP...
19:59:33 : LBA 41 849 452 try REMAP... complete
19:59:34 : LBA 41 849 453 try REMAP...
19:59:35 : LBA 41 849 453 try REMAP... complete
19:59:36 : LBA 41 849 454 try REMAP...
19:59:36 : LBA 41 849 454 try REMAP... complete
19:59:37 : LBA 41 849 455 try REMAP...
19:59:38 : LBA 41 849 455 try REMAP... complete
19:59:55 : Command interrupted by user!
19:59:55 : *** Scan results: Warnings - 0, errors - 8. Last block at 45 935 216 (24 GB), time 45 seconds.
19:59:55 : 8 of 8 defects successfully remapped.

Now I am left with i think one stubborn sector that I cannot in any way get rid of. It appears and disappears at random (Pending sectors value oscillates between 0 and 1 and offline uncorrectable is a constant one, despite the result of offline scan telling it was successfuly completed). Short self-test passes when the value is 0 and fails when it is 1. Current S.M.A.R.T. is as follows (after doing a Victoria remap):

Spoiler

WDC WD10JPCX-24UE4T0 WD-WXW1A17AP2FP
-----------------------------------------------------------------------------------
ID Name Value Worst Tresh Raw Health
-----------------------------------------------------------------------------------
1 Raw read error rate 200 200 51 949 •••••
3 Spin-up time 180 176 21 1958 •••••
4 Number of spin-up times 97 97 0 3268 ••••
5 Reallocated sector count 200 200 140 0 •••••
7 Seek error rate 100 253 51 0 •••••
9 Power-on time 96 96 0 3432 ••••
10 Spin-up retries 100 100 0 0 •••••
11 Recalibration retries 100 100 0 0 •••••
12 Power cycle count 98 98 0 2366 ••••
192 Power-off retract count 200 200 0 116 •••••
193 Load/unload cycle count 199 199 0 3836 •••••
194 HDA Temperature 111 98 0 36°C/97°F ••••
196 Reallocated event count 200 200 0 0 •••••
197 Current pending sectors 200 200 0 0 ••••• -this one keeps appearing and disappearing
198 Offline uncorrectable sectors count 200 200 0 1 ••••• -this one is always present
199 Ultra DMA CRC errors 200 200 0 0 •••••
200 Multi zone error rate 200 200 0 3 •••••
240 Head flying hours 96 96 0 3377 ••••

Ocasionally, during the scan, the drive hangs at the damaged sector, throws an error and then continues as nothing happened. (Sometimes it even spins down and starts back up after a few seconds after encountering the bad sector, sometimes it does not). This increases the pending sector count. Doing a scan with remap results in 8 consecutive sectors being remapped, and a normal scan after that, does not throw an error and the pending sector count returns to 0.

From what I know, the remap should occur when writing data (the drive writes to a spare sector instead of a damaged one). With that knowledge, I created a 32 gigabyte NTFS partition which covers the whole damaged area and tried filling it with data in various ways, however this did not help. Running chkdsk /r it finds and adds one sector to the Bad Clusters File and the pending sector again increases to 1. Running chkdsk again immediately after that returns that no errors were found.

Trying the h2testw tool, which is normally used for checking fake flash drives, it fills the whole partition with test files without hanging or any other issues, but throws an CRC error when verifying.
The test, however, finishes without errors after running chkdsk.

Since I could not remap in any way from within Windows, I decided to put the drive in a laptop, which supports legacy and has IDE mode of SATA controller in order to run MHDD and to try repair the drive. This software runs from CD and i think under DOS and is very low-level, but requires booting in legacy and does not work in AHCI mode.

I performed the full scan, which again delivered the same results and then, according to the procedure, first did a full drive erase (zero-fill) and then another scan with remap option turned on. However I ran into a problem, because for some time I thought the program hung when trying to remap (it is supposed to put a blue square next to the remapped block and it was not happening and the program was not responding at all and i could ony forcibly reboot). Turns out I just needed to be patient, because after couple minutes the blue square actully appeared indicating a successful remap. Waiting to the end it found and remapped 8 sectors and I thought this was going to be the end of it, but no. The normal scan still throws one error, and after 3rd try of remapping somehow 16 sectors got remapped, instead of 8.
I then tried in the opposite order (remap then zero-fill) and also tried erasing delays (this option basically zeroes any sector with access time over 350 miliseconds), but still, no luck, the stupid sector is still here.

I'm at a total loss here. What on earth happening? Why I cannot neither force the remap, nor does the drive do it by itself when it clearly should? Every other drive I encountered with couple bad sectors I could fix by the erase-then-remap method apart from this one.

Is there anything else I could try or should I just:
a) Create a partition and then chkdsk it, so that Windows is aware of the bad sector and from what I understand, will simply ignore it (h2testw seems to confirm it)

b) Leave the first 22 GB or so of the drive unallocated and create a partition covering the fully healthy area.

I would rather remap this sector since I don't want her to be scared by warnings that may pop up since there is this one sector pending... sometimes.

PS Is there any way I can forcibly set the APM to 254 or at least 128, or better yet, disable it? All the laptop WD Blue drives I dealt with have this value at 96 and this is gonna make the load/unload and spinup paramaters skyrocket since the drive spins down just after a couple seconds of no activity. When used as a system drive there is constant activity so it does stay running, but as a data drive it is not going to be accessed much. Is this aggressive spindown some sort of planned failure mode? Some toshiba drives have the APM set to 1 which makes the spindown happen every couple seconds, ridiculous ! I can set the parameter to 254 but it keeps resetting after each spindown.

 

[fzabkar]

The raw Read Error Rate is non-zero, so this drive should probably be retired.

 

[USAFRet]

Basically I got a 1 TB WD Blue 2,5 inch HDD from my friend to take a look at. The exact model is WDC WD10JPCX-24UE4T0 - nothing fancy, sadly it's only 5400 RPM.

Is (or was) there any life critical data on this drive?

Something that might lead to loss of job, divorce, jail?

If not, it is probably better to just let it go.
Given its current state, no matter what you do with it, it is not to be trusted.

 

[overclocler14]

The question is, is all your effort in remapping a marginal HDD (that can fail at any time) worth $10?

My goal is to try to do as much as possible without spending money. I OBVIOUSLY know that I can buy a new drive, but I think I did not stress this enough, but this drive is going to have only games on it, no critical data, so no matter if it fails. Game saves will be stored elsewhere so drive failure is not an issue ! Also it is going to be monitored every 2 months or so to check if more problems arise, but I doubt more bad sectors will appear since the remaining 99,9999% or so of the drive surface is in PERFECT condtition and it does not have any slowdowns or other issues and the read/write speeds are fine.
I am totally aware this one has some issues but on the other hand is just one sector, not hundreds.

Do you really want me to toss a still-working drive just because of a single bad sector?
I have a 1 TB 3,5 inch Seagate in my PC as a main gaming drive and one day it randomly threw 128 uncorrectable sectors, so I panicked and bought a new one, fearing it's gonna die any day. It turned out it was some kind of random logic error. Over 2 years later it's still going strong and currently has ZERO ISSUES and has decent (for HDD anyway) sequential read/write speeds of around 200 MB/s. The only thing I had to do was to transfer my games to a bunch of different drives I had laying around and to do a zero-fill and all of the Pending and Offline UNC disappeared from SMART, no remaps were done, and all the surface scans pass perfectly to this day. I check the drive every month or so, when I have time and nothing abnormal is happening.
But according to your recommendations I should have got rid of this drive and buy a new one...

My question was regarding the other methods I could try to make the drive ignore this single sector (as it is meant to do normally), not whether I should keep using it or whether it safe to store data on it. I once again repeat that me and my friend are completely aware that this HDD is not perfect and no precious data will be stored on it, but if all it takes to get it going is to remap a single sector, then yes, I am going to bother and spend time trying.

The raw Read Error Rate is non-zero, so this drive should probably be retired.

None of the diagnostic programs (even the WD one) report that this is an issue. In fact, even with the one pending sector, the WD data Lifeguard says that all drive parameters are within tolerance. Also, noticed that all Seagate drives have this number immediately skyrocket even on a new drive and I noticed it is the exact same value as ECC errors recovered, it is always high and fluctuates a lot.

https://ebay.com/itm/174589221142 $10, 1TB 2.5" HDD enterprise-class, used w/ 30-day warr, use USB adapter

https://ebay.com/itm/156189458646 ~$15 1TB 2.5" int HDD open box

The question is, is all your effort in remapping a marginal HDD (that can fail at any time) worth $10?

I have no idea why but for this "enterprise drive" the shipping appears to cost me over $1000 (?), anyway it is shipped from the US so does not make sense, since I am in Europe (Poland), and I can source a drive here.

The second one is probably gonna be worse than the one I currently have since it is a freaking STB/surveillance HDD with SATA II interface...

I wouldn't consider this drive "marginal". For me, marginal is the one pulled from an old 2009 Toshiba laptop, it has more than a couple hundred of sectors w/ access time over 150 ms and dozens of remaps, and it is insanely slow towards the end of the surface scan (I'm talking speeds of less than 10 MB/s, sequential read). This one I consider to be indeed gone, however, since it still somewhat works so I use it as a "data dumpster" where I put completely irellevant stuff I do not care to lose at all.

 

[fzabkar]

None of the diagnostic programs (even the WD one) report that this is an issue. In fact, even with the one pending sector, the WD data Lifeguard says that all drive parameters are within tolerance. Also, noticed that all Seagate drives have this number immediately skyrocket even on a new drive and I noticed it is the exact same value as ECC errors recovered, it is always high and fluctuates a lot.

Data Lifeguard is designed to minimise warranty returns. To this end it will give a marginal drive a passing grade. It is not to be trusted.

Seagate's SMART attributes are counter-intuitive. Seagate's raw "error rates" are actually sector counts or seek counts, so they "skyrocket" even for perfectly healthy drives. WD's SMART attributes, OTOH, record an actual error count in the same position.

Seagate's Seek Error Rate, Raw Read Error Rate, and Hardware ECC Recovered SMART attributes:
https://www.hddoracle.com/viewtopic.php?f=6&t=89

Edit:

If you are interested, I can show you how to retrieve several potentially embarrassing SMART attributes that WD hides from the user.

 

[overclocler14]

So I'm back with progress. I successfully removed the one stubborn Offline uncorrectable sector (198) by plugging the drive into an external SATA power source and leaving it overnight. The drive was not connected to any other device, so it was idle for hours and i think successfully completed it's off-line tests and updated the value.

I then created a 22 gigabyte partition followed by the next one which filled the rest of the drive and then removed the first partition, so that this space is unallocated and cannot be used by the OS. I think I'll live with losing a bit of capacity in exchange for higher reliability.

I then ran chkdsk on the remaining partition and it came out without any issues and also all of the important SMART values remain at 0 since there are no attempts of reading in damaged area.

I think this case is solved for now and the drive is ready for use. Should other issues come up they will be detected in the scan.