xfs_repair lead to 2.4T of files in lost+found

[NIronwolf]

I'm hoping someone has any sliver of decent news about my situation.

I had a drive halt and it stopped the array.  When I rebooted it came up as a bad file system so I stopped it and restarted in maintenance mode so that I could run the xfs_repair check from it's settings page. The repair finished and I was then able to get the array to start up again.  I set about seeing what had happened on the drive and now I'm faced with 10s of thousands of numbered files in my 'lost+found' directory.  There are a few directories, but a huge amount of this drive was various appdata subdirectories and that means I have no idea where these files belong even when I have a directory listing of filenames.

I guess I'm just looking for confirmation that I'm completely hosed.

Thanks.

[itimpi]

In most cases sorting out a lost+found folder is more effort than is worth it.   The items in there are ones where the directory information giving their names is what was lost.  Sometime using the linux 'file' command to determine what type of content is in a file can help - but in your case it sounds as if that might not be much help.

 

It just emphasises that you still need backups of anything that you deem important as things can go wrong that are outside your control.

 

Just as a check are you saying that the drive is NOT disabled (i.e. has a red 'x') against it.  Just asking as if it was disabled the repair would have run against the emulated drive, and if that is what happened a repair of the physical drive might be more successful.

[NIronwolf]

I'm not positive at this point.  It may have been marked disabled when started in maintenance.  I've been having issues with my external port replicating enclosures so I may just be headed for a big crash.  I'm finding some bulky files with recognizable names.  I'm hoping that I can find a decent backup of my appdata stuff, but I usually take that on the 1st of the month so...  yeah.  And ofc it's on the array instead of a separate system so worried about the reliability of that.

[itimpi]

33 minutes ago, NIronwolf said:

nd ofc it's on the array instead of a separate system so worried about the reliability of that.

If the appdata share is also on the array then make sure they are configured to use separate drives.   Each disk in the main array is a self-contained filing system so you would be unlucky to get a failure that caused simultaneous corruption of multiple drives.