Chkdsk says over 400 GB is in bad sectors - could it be wrong?

[smirk]

Hi, I have two 2TB WD20EARS hard drives installed in a home server. Last night I ran Chkdsk on them, and this was reported:

Disk 1: "170650768 KB in bad sectors."
Disk 2: "414378904 KB in bad sectors."

I pulled one of the drives, put it into another PC, installed Western Digital's Data Lifeguard tools, and ran an extended scan on it. In the morning, it finished and said everything passed. In addition, if you examine the SMART data, everything has green checkmarks and there are no warnings. The Reallocated Sector Count SMART entry shows a value of 200 on both drives and raw data of zero, so I'm guessing that 200 is the default value (but I honestly don't really know how to interpret SMART values). However, running Chkdsk again reports that high number of bad sectors.

Can anyone comment on what "xxx KB in bad sectors" means? Does it mean that my data resides on bad sectors and is damaged? Does it mean that those sectors have been flagged as bad and the data has been moved to spare sectors? And why does Western Digital's utility show no problems, even after doing a surface scan?

Please help, I'm a little bit freaking out right now. All my data is stored redundantly on these two drives, and I can't tell if a problem actually exists or not.

Thanks!

 

[imagoon]

It is how many sectors are marked bad in the file table. If there was something causing read errors [like an overclock] you can get bad sectors marked. If the drive tools passed it, I would expect it passed. You may need wipe the disk with the tools to recover it.

 

[smirk]

Thanks for your quick reply.

Are these bad sectors marked at the OS or drive firmware level? Because SMART shows no bad sectors that I can see. And more importantly, is there any way to tell if any data has been lost?

(And no overclock, it's a plain vanilla setup. I suppose it could be bad SATA cables, but it seems odd that two different cables would both go bad at the same time.)

 

[imagoon]

File table is "OS" level. You would need to scan your disk and see if files reappear as filexxxx.chks. I would be looking in the event log for bad sector notes to see what the error code was. It might give you an idea about what is going on.

 

[razel]

As long as you did the full surface scan with the WD Diag and it didn't report anything else, then it's time to update the file system. You can do this with chkdsk. Assuming the file system is NTFS and you are running Vista/7 at a command line run:

chkdsk [drive letter]: /V/B/X

/B Re-evaluates bad clusters
/V verbose output
/X force dismount (the drive will be unavailable in Windows Explorer)

 

[smirk]

As long as you did the full surface scan with the WD Diag and it didn't report anything else, then it's time to update the file system. You can do this with chkdsk. Assuming the file system is NTFS and you are running Vista/7 at a command line run:

chkdsk [drive letter]: /V/B/X

Thank you, I will try that. I did do a full surface scan with the WD Diag and it merely said the drive "passed" (this is not a wordy app).

However, I may have made the mistake of connecting the drive to an XP system and running "chkdsk F: /R". I'm afraid to abort it now, but hopefully it won't harm things by trying to repair the disk, when all that really needs to be done is have a Windows 7 system recheck the sectors by using "chkdsk /B".

 

[Cerb]

Thank you, I will try that. I did do a full surface scan with the WD Diag and it merely said the drive "passed" (this is not a wordy app).

However, I may have made the mistake of connecting the drive to an XP system and running "chkdsk F: /R". I'm afraid to abort it now, but hopefully it won't harm things by trying to repair the disk, when all that really needs to be done is have a Windows 7 system recheck the sectors by using "chkdsk /B".

XP typically removes, or at least makes FUBAR, search indices, IME. Nothing to worry over.

 

[smirk]

As long as you did the full surface scan with the WD Diag and it didn't report anything else, then it's time to update the file system. You can do this with chkdsk. Assuming the file system is NTFS and you are running Vista/7 at a command line run:

chkdsk [drive letter]: /V/B/X

You are the KING! Thank you, that did it!! Zero bad sectors on both drives now.

Which kind of begs the question of how did this happen in the first place? I pored over the event logs and found two errors by "PlugPlayManager" that said a drive was disconnected without properly ejecting, and that corruption may occur. Since these are internal drives, I know they weren't disconnected -- so what could cause this kind of thing?

 

[Binky]

This seems like a nice little scare to remind you to back everything up.

 

[Aikouka]

Interestingly enough, my WD20EARS is also acting up on me. I don't know if you were having issues with files, but there are files that I cannot even access without hanging up Windows Explorer.

Some of the stuff in here might be interesting to try and see what happens. I've also got one WD 1.5TB Caviar Black that's acting up on me too. There are two files on it that I cannot touch without hanging Explorer (and being unable to close it without rebooting).

 

[smirk]

This seems like a nice little scare to remind you to back everything up.

Ironically, these two drives make up the storage pool in my Windows Home Server machine -- this *is* my backup! But you're right, I need to make backups of the backups more frequently.

 

[smirk]

Interestingly enough, my WD20EARS is also acting up on me. I don't know if you were having issues with files, but there are files that I cannot even access without hanging up Windows Explorer.

Hmm, that's interesting. The OS in this rig is Windows Home Server, which generates log entries that I don't really understand. I do see a lot of messages like this:

***ERROR: DEfileFromShadow(c:\fs\D\DE\shares\Music\Depeche Mode - Sounds of the Universe\01-In Chains.mp3,f07aadb6-2411-4e7f-940d-284aab86f69e) took too long. 580.9491ms

"Took too long"? No idea what that means or how serious it is, or if it's similar to what you're experiencing, or if it means the drive is on its way out (even though SMART says it's fine).

???

 

[Aikouka]

Most of my computers (including my "server") use Windows 7 Professional. I tend to get entries in my Event Viewer about disk errors.

My main machine has been reporting disk errors for Disk6. I have 4 HDDs that are connected via SATA internally (one other via USB 3). That would make me assume that they are Disk0-3. I have a few network drives, and the one that was causing problems should be #3 (D,E,F,G,H,I).

My server has been reporting similar Disk errors. I'll have to verify that they have gone away as I recently replaced a 2TB HDD with a 3TB HDD, and I then moved any salvageable data off the WD drive onto the recently replaced 2TB HDD.

I did try to play a music file off of the "newer" 2TB HDD, and it was rather warped/garbled. Even Robocopy had issues trying to get stuff off that drive, so I'm not surprised.

 

[Binky]

I had some odd smart errors on WHS v1 that turned out to be related to an outdated Intel RST that seemed to be conflicting with some newer WHS update, or another driver update. Mine would suddenly show a drive as failed on every single smart measure. This does not sound anything like your issue(s), but it might be worth looking into if your WHS uses RST.

 

[smirk]

I had some odd smart errors on WHS v1 that turned out to be related to an outdated Intel RST that seemed to be conflicting with some newer WHS update, or another driver update. Mine would suddenly show a drive as failed on every single smart measure. This does not sound anything like your issue(s), but it might be worth looking into if your WHS uses RST.

Thanks Binky. I'm not using RAID and the drives aren't using AHCI so that's probably not it, but it's good to know about your experience.

 

[taltamir]

Thanks Binky. I'm not using RAID and the drives aren't using AHCI so that's probably not it, but it's good to know about your experience.

RST is not just RAID drivers. RST is "Rapid Storage Technology".
If you are using AHCI or even IDE mode you still need to have the latest intel RST.

 

[Cerb]

RST is not just RAID drivers. RST is "Rapid Storage Technology".
If you are using AHCI or even IDE mode you still need to have the latest intel RST.

No, you don't. Unless you use special features that MS' drivers don't support, RST is not needed.

 

[taltamir]

No, you don't. Unless you use special features that MS' drivers don't support, RST is not needed.

There are always bug fixes performance improvements included in newer drivers
http://downloadmirror.intel.com/20624/eng/iata_10.8.0.1003_ReleaseNotes.htm
http://www.station-drivers.com/telechargement/intel/sata/11.5.htm

While most deal with RAID issues, not all are RAID issues. There are various non RAID related bugs that are fixed.

 

[Cerb]

There are always bug fixes performance improvements included in newer drivers
http://downloadmirror.intel.com/20624/eng/iata_10.8.0.1003_ReleaseNotes.htm
http://www.station-drivers.com/telechargement/intel/sata/11.5.htm

While most deal with RAID issues, not all are RAID issues. There are various non RAID related bugs that are fixed.

I saw all of a single fixed issue that wasn't RAID, or a bug of RST, there, and it's not one I'd worry much about (an ODD that's off not showing up in device manager). The only other issue, also for ODDs, hasn't been fixed, and reads like it's pretty rare, as well (eject lever?). Bug fixes to RST are rather useless bug fixes, if you don't use RST.

 

[taltamir]

I saw all of a single fixed issue that wasn't RAID, or a bug of RST, there, and it's not one I'd worry much about (an ODD that's off not showing up in device manager). The only other issue, also for ODDs, hasn't been fixed, and reads like it's pretty rare, as well (eject lever?). Bug fixes to RST are rather useless bug fixes, if you don't use RST.

well, they are fairly mature drivers from a company that prides itself on quality control. And the release notes I linked were for a small incremental update. I'd be really concerned if they had many big bugs to fix with drivers that have been around for so long. I have been having difficulty finding the release notes for older versions though to show what bugs those had.

 

[Cerb]

I'm not bashing RST. You just don't need them to have perfectly functional and performant plain old storage volumes, provided the features you need are handled by MS' driver (FI, an old ICH w/o AHCI may need RST to enable hot-swap and NCQ; but most boards do it fine w/ MSACHI).

 

[smirk]

Thanks guys. It sounds like RST couldn't hurt, so I may as well install the driver to see if my system benefits.

Since Windows Home Server isn't as widely used as, say, XP is I will probably never know what those "file access took too long" messages mean, but we will see if they continue with some new drivers.

 

[Binky]

The issue I mentioned above was an outdated driver in a system that came with RST preinstalled. I'm not suggesting that you install RST if your system was running without it...

 

[taltamir]

I'm not bashing RST. You just don't need them to have perfectly functional and performant plain old storage volumes, provided the features you need are handled by MS' driver (FI, an old ICH w/o AHCI may need RST to enable hot-swap and NCQ; but most boards do it fine w/ MSACHI).

You are correct.
When I said that you "need" it I was using a word that is too strong and thus incorrect.
It is essential for running RAID, it is compatible with AHCI, but you can do just fine with MS built in AHCI driver.

 

[razel]

Hmm, that's interesting. The OS in this rig is Windows Home Server, which generates log entries that I don't really understand. I do see a lot of messages like this:

***ERROR: DEfileFromShadow(c:\fs\D\DE\shares\Music\Depeche Mode - Sounds of the Universe\01-In Chains.mp3,f07aadb6-2411-4e7f-940d-284aab86f69e) took too long. 580.9491ms

I wouldn't worry about it. I'm assuming the word Shadow relates to Shadow Copies, which I freaking LOVE on servers. You can check by right clicking to the properties of a folder or file and look for a 'previous versions' or 'shadow copies' tab. You should then see a list of dates that you can chronologically go back and revert files or folders to.

The 'Shadow Copies' tech is also used when doing backups. Obviously when you start a backup and you write to a file, it will make parallel copies. The updated one on your disk and the old one somewhere in another universe.

Your drives are advanced format so, I'd just double check to see if your drive is aligned and has allocation unit sizes of at least 4k.

- You can check alignment with 'msinfo32' look under Components : Storage : Disks. Then look for 'Partition Starting Offset' It should be evenly divisible by 4096.
- You can check allocation unit sizes with chkdsk. It should say it after it's done running.

If your drive isn't aligned WD does provide WDAlign to use.