Question Is occasionally removing an external drive without ejecting cause for file system corruption? Or just a crappy chipset on adapter?

[viivo]

I use a 2.5" SSD in an external USB enclosure for storing various backups, and twice now I have lost the contents because the file system became corrupted. Is this due solely to the fact that I sometimes unplugged the USB cable before fully ejecting it, or could something else be causing it? I'd rather not take a chance on losing my files again if, for example, the chipset (Jmicron JMS567) of the USB adapter is to blame and not the premature non-ejection.

 

[pcgeek11]

Yes, it could and frequently does cause file corruption by pulling the drive without ejecting the drive normally.

Stop doing that. It only takes doing it once at the wrong time.

 

[Shmee]

Don't just yank it out, if at all possible. That said, the issue could be the SSD or the adapter.

You could try a SATA connection to see if that helps for viewing/recovering the files, and even run a recovery program, if it is still corrupted. You could still get your data back.

 

[viivo]

I tried a few different recovery programs and neither could retrieve anything, even the 5+ hour deep search functions. I don't know if that means lots of bad clusters and the drive is on its way out or if the file system corruption just did a good job of screwing over my files.

Thanks for the replies.

 

[Shmee]

What model SSD is it, and how old? You should at least check its SMART.

 

[UsandThem]

You can also see how the drive is configured in the device list.

If you don't always eject it through Windows first (which you really should), you can disable write caching so the drive can be better protected from sudden power loss. Also, like Shmee mentioned, you can run the manufacturer's utility to see the health of the drive (or a utility like CrystalDiskInfo).

 

[viivo]

What model SSD is it, and how old? You should at least check its SMART.

Crucial MX300. I've had it for some time, but it's seen relatively little use. The only backups I stored on it are small configuration files, documents, etc.

You can also see how the drive is configured in the device list.

If you don't always eject it through Windows first (which you really should), you can disable write caching so the drive can be better protected from sudden power loss. Also, like Shmee mentioned, you can run the manufacturer's utility to see the health of the drive (or a utility like CrystalDiskInfo).

For sure, I always make sure that's not checked.

Despite the horrible things I've read about the Jmicron chip used on this, the only way to know for sure is to try using it again and ejecting every single time. Hopefully that's all it was. Thanks again for the responses.

 

[damian101]

Only if you use a file system without journaling, like FAT32 and exFAT.
You can fix corrupted FAT32 and exFAT partitions in Windows with "chkdsk <insert drive letter>: /r". And even if your file system gets irreparably damaged you are very likely to still be able to access the data with a recovery tool like DMDE.

 

[MalVeauX]

Heya,

It's a pretty bad idea to pull a drive that is mounted with a file system that cannot or simply does not compare source/destination data to ensure integrity. Some files systems can handle this. But none of them are Windows file systems (unless spinning in virtualization on a superior file system running behind it, like ZFS).

Very best,

 

[damian101]

@MalVeauX There is ReFS on Windows, it's a copy-on-write filesystem like ZFS and Btrfs, and like them it also stores checksum metadata to protect against silent data corruption.

 

[MalVeauX]

@MalVeauX There is ReFS on Windows, it's a copy-on-write filesystem like ZFS and Btrfs, and like them it also stores checksum metadata to protect against silent data corruption.

That is true, I digress, I worded that poorly and disregarded REFS and unfairly left it out. But REFS merely tries to compete with ZFS's superiority with respect to data integrity and features and overall robustness, but it's only a valiant effort frankly. But you are right, REFS does technically try.

Very best,