I ran a [correcting] parity check recently and it fixed up 13 parity sync errors. I wasn't expecting this and as a result, ran another one the next night [non-correcting] which picked up another 9 parity sync errors. I'm not sure whether this is related to my recent migration of disks to XFS (see below) or if I have a disk on the way out.

 

There aren't any SMART warnings that I'm concerned about at this stage, I did a zero-byte search against my shares and found two TV eps with zero bytes but this might be unrelated. On another share however (different disk set) I have 227 zero-byte files. Some are expected, random html and config placeholders but there is some music and old documents which wouldn't be expected. Almost all of these appear to originate on Disk 2, so I'm running a full SMART check on that disk now. The disk is pretty old and so is the data, could this be bitrot?

 

Any suggestions on what might be causing this? If the disk checks OK, I'm wondering if a new config might be in order. Just worried about potential data loss at this stage.

 

Background relating the filesystem changes:

I recently migrated a bunch of disks from ResierFS to XFS following what guides I could find, I wasn't overly concerned about maintaining perfect metadata or even parity during the moves, just that the files moved OK. I used a Krusader docker to shuffle the data on to a blank XFS formatted disk and then changed the FS on the old, now blank Reiser disk to XFS, let unRAID format the blank disk then ran a parity check. With each move there were about 12/13 parity errors corrected which I attributed to the shuffling of ~2-3TB of data and wasn't overly concerned but now I'm wondering if I should have been.

unraid-diagnostics-20190811-0100.zip

Disk 2 passed an extended SMART check.

If I were you, I would install the Dynamix File Integrity plugin to watch for corrupted files. That's an easier way to do it. If there's parity error but all the files seem ok then it might very well be the parity disk. Maybe it wasn't built correctly?

You should also run memtest.