I have Redhat 8 running on hosted server - it's been running great for years.

All of a sudden, last night, almost every file I try to access comes back with "input/output error". Some directories, most notably /usr/bin - are completely unreadable. Others will allow ls to display the file size and permissions, but when I try to read the file, the same input/output errors occur.

I assume this is a massive hard drive failure? Could 80% of the blocks have gone bad simultaneously? I'm no Linux expert, so I want to make sure that I'm not missing something obvious.

The problem is, there is some database data (in a nightly export file) that I desperately need to access.

I asked the hosting company to mount it as secondary drive. Should I try running disk utilities on it? Or will that risk causing further damage? Should I just send the drive to a professional recovery service?

Thanks for any help - I'm in a real bind.

You don't have to have 80% of the blocks go bad. Just enough to corrupt the file table. So yeah, it's possible.

As an FYI, I've used hard drive recovery services at work on failed drives. Anecdotally we've paid between $2,000 and $8,000 for drive repair.

Using fsck with the -N option should prevent it from making any changes. There is also a -n option, though I'm not sure what the difference is between the two. I would verify that fsck for your fileystem has the -N option and give it a shot (probably with -V too).

If I was your mother, I'd be telling you that you should've backed up like I told you too.

Good luck.
-fudam