Windows 10: Hard Drive Bad Sectors found by WD Data Lifeguard but not by chkdsk

Hi all

Here's the story... I had 2x 3TB HDD's in my NAS drive that was reporting that one of them had bad sectors, this was confirmed by Western Digital's HDD Utility tool, so I replaced it.

I've reformatted it and hooked it up via a USB caddy and ran chkdsk on it but that never found any problems with the drive and it still seems to work OK, copying large & small test files back & too. I've once again ran the WD utility and it passes the quick test, however it fails on the extended test (& cannot repair the bad sectors it found when it attempted to)

How come chkdsk isn't finding any errors? I was hoping it would & would then map the bad sectors and stop Windows from attempting to put any data there, and I could continue to use it for data storage, backups etc...

Any thoughts?

Thanks in advance...

:)

 

File Missing After CHKDISK

You can find the manual for the WD Data Lifeguard Diagnostic here: http://products.wdc.com/support/kb.ashx?id=WFZgCB

The error codes are listed on the bottom. Which test did you perform (quick or extended)?

If the tool found some bad or unreadable sectors you might be able to run chkdsk /r from the cmd to try to isolate those bad sectors and transfer the data from them to other healthy sectors, but there's a chance of data loss. Have you tried the tests with a
different cable on a diffrent port or on a different system?

.

 

Unable to access external Hdd under Windows 10, "Correcting errors in the uppercase file"

Ahoy there Captain.

I've successfully removed the data from the drive and have now wiped it and ran the data lifeguard tool. Data lifeguard fails to complete the test due to there being too many bad sectors. I bought the drive in 2010 so I've had a good run, it's had a lot
of use. In fact this is the first WD drive that has failed on me in nigh on 10 years, so I can't really complain. Thank you for your help.

 

If its still under warranty send it back. if not junk it. Why take any chance of losing data hard drives are too cheap these days to take any chances with them. Typically once a hard drive starts dropping sectors it will continue dropping them.

 

The bad sectors will have been marked as bad so chkdsk wouldn't test them

 

It's (typically.. *Rolleyes) about 4months out of warranty if I remember correctly, it was over £100 so still a fair bit of cash, also it seems a shame to bin it if it might not be totally trashed. As you say though, once they start developing bad sectors it might be the beginning of the end for it. I'm gonna keep it hooked up and test it every day or two - it's empty now so I won't lose anything...

Ah, right. What marks the bad sectors...the SMART function of the drive, or the scan by the WD utility?

 

Bad sector mapping is done internally by the drive but can be triggered by external events.

When an error occurs when a sector is written it will be immediately replaced. This is done internally by the drive and even the OS is not aware that the error occurred. The problem is that there are only a limited number of spare sectors available and when that number is reached no more errors can be repaired. That number is typically not documented.

Things are more complicated when a read error occurs. In this case the location of the sector is noted by the drive but it is not replaced. Doing so would eliminate any possibility of recovering the data in the sector. There are 2 ways the sector may be replaced.

1. If the sector is later read successfully (the problem may have been temporary) it will be replaced with no data loss.

2. The sector is written to. The data is then known and the sector can be safely replaced with no data loss.

Unfortunately you have no assurance either event will ever occur.

I don't know exactly what the WD software does. It may be using drive specific commands.

Typically the cause of drive errors is unknowable. But once errors start to accumulate the situation often gets worse over time. You might have a few errors today and thousands next month, or next week, or tomorrow. No way of knowing.

 

Did you use the /R option with chkdsk? It won't look for bad sectors unless you do. WARNING: chkdsk /r can take a very long time to complete - we're talking hours here, tens of hours if there are many bad sectors.

The real warning signs of a failing drive are to be found in its SMART data. I use CrystalDisk Info to look at that, preferring the Portable App version (no install required). https://portableapps.com/apps/utilit...kinfo_portable

The key warning signs to look for are...

https://www.backblaze.com/blog/what-...rive-failures/

When any one of these is greater than zero, start to worry, if it's more than a handful, worry a lot. Two or more then panic, especially if they are increasing with time.

 

Thanks for the really informative posts....*Wink*Smile
@Bree I download CrystalDiskInfo s/w you recomended and out of the 5 drives in my PC, sure enough, the 3TB one had a yellow 'Warning' flagged up, with the culprit, as you can see from the screen caps - being the '05 Reallocated Sector Count' . What I'm a bit confused about though, is the other drives, which were blue 'Good' flags, ALL had a higher reading than the one flagged up on this score...but classed as good??

 

There are some tests that are common on the software.
Some of the disk testing software have different tests and may detect problems not found using another software.
The chkdsk tests depend on which switches are used.
It tests most of the disk but is not testing the whole disk.
When running the tests find software that will test or report each: Health (SMART), benchmark, full error (surface) scan, long generic, etc.
These are commonly used software to test disks: HD Tune, HD Sentinel, Crystal Disk, Sea Tools for Windows
For problematic drives you may be able to use them for very long periods of time even with bad sectors and many continue to use drives until they fail. Others will replace disks at the first sign of a problem.
Performance may suffer on a drive without having failure and this may be another reason to replace before failure.
If you make a backup image using Macrium you will indirectly test the drive file system.
And with a backup image you may be able to use the drive until it fails.

 

The WD scan will be more thorough than the Windows version. So it's not unexpected to be different.
You have found that out.

Once bad sectors start to be marked off statistically it's a bad sign, if they increase over time an impending failure.

I have had HDDs mark off a few bad sectors and they have never changed and subsequently lasted a very long time. Others have failed completely without warning due the electronics dying on the drive.

SMART data is not a reliable guide to impending failure, studies have shown that it's more of a rough indicator on a proportion.
Different HDDs use different technologies, so there is going to be a variation on what is reliable guide data to an impending failure.

 

SMART warnings can be compared to the check engine light in a car. In both cases the warning needs to be investigated. Both warnings are useful but you should not rely on them. The lack of a warning is no assurance that all is well. Predicting component failure has always been very problematic and science has a long way to go in this regard. A component may be near failure but not yet effect functionality and all tests will show no problems.

Any hard drive, new or old, may fail at any time, often without warning or apparent reason. I had one drive that showed no warnings of any kind. The next day it wasn't even recognized by the BIOS. As all important data was backed up I did not investigate beyond checking cables, etc.

SSDs, like most electronic components, typically fail without warning.

 

It's the Raw Values you should look at, these are the actual count of reallocated sectors. For the 3TB 'bad' disk there were 0x46 (70 decimal). The good disk has zero. The 3TB's Read error rate looks worrying too, that should also be zero.

I wouldn't trust this disk for archive purposes, it could degrade at any time.