FAT32 drive repair question

[weemadando]

I've got a WD external drive that's formatted to FAT32 and it's been fritzing on me - giving me read errors occasionally and causing crashes when I try to read from some folders.

What tools/software would people recommend I try in order to maintain/repair this drive without having to format it and lose a shitload of important data?

[Bounty]

For starters, get your data off that drive ASAP.

[Fingolfin_Noldor]

Check your cables. USB Cables have a knack for breaking after rough use.

[phongn]

Pull data off the drive immediately and replace it. If it's failing there's nothing you can do to maintain it, repairing it might exacerbate the damage and FAT32 does you no favors here.

[Dominus Atheos]

There's no need to replace it just yet. You absolutely need to pull your data, but then you should reformat the disk as NTFS and run the long S.M.A.R.T. test in HD Tune. If that passes, that means the drive is fine but if it fails then it needs to be replaced.

[phongn]

Drives can fail silently; a non-failure result may be meaningless.

[weemadando]

Alright. Data transfer ahoy. To be following shortly by formatting to a file system that doesn't fucking suck.

Or do people think that this is a problem with the drive itself?

[Dominus Atheos]

Drives can fail silently; a non-failure result may be meaningless.

An extended S.M.A.R.T test checks things like the temperature of the bearings in the drive, the vibration of the drive caused by the spinning of the disk, and the readability of every sector on the drive over the course of several hours of heavy use. There should be next to no way the drive can be failing and the extended SMART test doesn't pick it up.

[weemadando]

All attempts to back up any data are resulting in a Windows Explorer lockup.

Reformatting to NTFS. Goodbye all my precious, precious work.

[Uraniun235]

Drives can fail silently; a non-failure result may be meaningless.

An extended S.M.A.R.T test checks things like the temperature of the bearings in the drive, the vibration of the drive caused by the spinning of the disk, and the readability of every sector on the drive over the course of several hours of heavy use. There should be next to no way the drive can be failing and the extended SMART test doesn't pick it up.

Unless, of course, the failing components weren't mechanical, but electronic - i.e. the chipset mounted on the underside of the drive.

[Stark]

All attempts to back up any data are resulting in a Windows Explorer lockup.

Reformatting to NTFS. Goodbye all my precious, precious work.

Precious work should have been backed up.

Don't Windows upgrades offer to 'convert' filesystems FAT32->NFTS anyway? I dimly recall that sort of thing in XP.

EDIT - It's probably too late, but I hope you tried to move your data in small bits; if you tried to drag everything at once and it failed, you can often just avoid the corrupt area and copy other stuff off.

[weemadando]

Yeah, I was able to salvage some bits and pieces. All of my really precious stuff was backed up with layers of redundancy. It's the gigs and gigs of affiliated stuff that I can't back up to a cloud server easily that I'm pissed about. Thankfully some associates took recent backups so I can recover some stuff that way.

[Dominus Atheos]

Drives can fail silently; a non-failure result may be meaningless.

An extended S.M.A.R.T test checks things like the temperature of the bearings in the drive, the vibration of the drive caused by the spinning of the disk, and the readability of every sector on the drive over the course of several hours of heavy use. There should be next to no way the drive can be failing and the extended SMART test doesn't pick it up.

Unless, of course, the failing components weren't mechanical, but electronic - i.e. the chipset mounted on the underside of the drive.

Like I said, it tests the readability of every sector of the drive many times over the course of several hours. (I just canceled the one running on my 1TB drive after 2 hours because it was only a 1/3rd of the way done. The drive was running at full bore the whole time) If there's an issue that's causing read errors, the extended test will pick it up. If it runs for 6 hours straight at full bore and doesn't have any errors, I'd be willing to stake money the drive doesn't have an issue.

More info here and here.

[Xon]

If this was an NTFS formated disk causing explorer to lockup, I would call it is faultly hardware in a heartbeat.

But FAT32 can trivially corrupt itself in ways which will cause application crashes even on NT based kernals. It shouldn't kernal panic/bsod however, but I wouldn't be sure.

FAT32 is insanely fragile, and on a external portable(just yanking it could cause random data loss) I could see it being corrupted quite trivially.

[Edward Yee]

I've got a WD external drive that's formatted to FAT32 and it's been fritzing on me - giving me read errors occasionally and causing crashes when I try to read from some folders.

Quick question... on my desktop with XP Pro, what if I have the crashes-when-trying-to-read from certain folders, and when selecting video files, but only in Windows Explorer? That is, only Windows Explorer crashes, but checking out those folders/picking those video files in any other 'browse' function (i.e. the Open menu of Windows Media Player) doesn't result in such an error.

[Marcus Aurelius]

If this was an NTFS formated disk causing explorer to lockup, I would call it is faultly hardware in a heartbeat.

But FAT32 can trivially corrupt itself in ways which will cause application crashes even on NT based kernals. It shouldn't kernal panic/bsod however, but I wouldn't be sure.

FAT32 is insanely fragile, and on a external portable(just yanking it could cause random data loss) I could see it being corrupted quite trivially.

It's not that FAT32 is insanely fragile (well, it is pretty fragile, but it does not explain this), but since it's not journaled in any way, you just can't expect it to work flawlessly if it's just yanked. Most of the time you can recover though if you fsck it (pun intended) immediatelly afterwards. You can fuck up even a reasonably modern FS (such as ext2 or non-journaled UFS) by yanking an external drive without flushing the cache if write caching is enabled. (Windows XP SP1 and later disable write caching to USB flash drives by default, by the way.)

As for BSOD, I don't think FS errors can do that unless the system files or startup partition are corrupted. Braindead USB hardware and drivers on the other hand can BSOD Windows, but it is not common with current USB implementations. However, if you have some of the older VIA USB chipsets...