- Minor
I have two SSDs in btrfs raid0. I've been fiddling around with my box lately, and I guess the cables were loose on one my drives, because the drive dropped while the mover was running. It came back after reboot and no data was lost, but there were some very preventable corruption issues.
Bug #1:
After one disk dropped, the files on the raid0 cache remained visible to the UI (but not readable through the UI). Good, I guess. But the mover kept on going and wrote corrupted files to the array disks which had the appearances of being functional files. I was able to clean them up because the files remained on the cache (presumably it went read only). But the mover should not do this!
Bug #2:
Unraid refuses to mount the cache if one drive in btrfs raid0 is not available after reboot. Good. And the drives sync after being reconnected and the system rebooted. Good. But even though it works and the data is good, unraid shows the dropped drive as unassigned, not as part of the cache pool. I can't see any way to add it back without moving everything off and wiping it. This is very inconvenient.