I was working on upgrading my parity disk and swapping smaller/old disks (with no data on them/all data migrated to other disks using unbalanced) with bigger new ones and used New Config to reset the array. Had a screenshot of my array layout so I knew what disks went where. Pretty sure I had done this a while ago with no issues, just start the array and do a parity check.

This go around when I tried to start the array, all of my data disks had an error like unrecognized or unmountable partition. Weird, they were all just working fine with an xfs file system (except the former parity disk I moved into the array which hasn't been formatted yet). Came across this post with a similar issue - https://forums.unraid.net/topic/192800-after-new-config-several-drives-are-not-recognizing-the-file-system/.

Manually selected xfs for all 8 of my data disks, started the array, and it started right up prompting me to format the 1 former parity disk. All my data/shares are good, no data loss. Of course I'll need to do a parity check. Playing with fire, I pulled one disk out of the array leaving the disk unassigned and that slot in the array empty. I could not mount the disk using unassigned devices, disk log gave me a unraid unassigned.devices: Partition '/dev/sdi1' does not have a file system and cannot be mounted. Did another New Config and put the disk back, selected xfs manually on all the data disks again. Array started fine, no data loss again.

JorgeB recommended doing a wipefs /dev/sdX1 in that other thread while the array is stopped, I did that to the disk I tried to mount with unassigned devices and got this:

root@unraid:~# wipefs /dev/sdi1

DEVICE OFFSET TYPE UUID LABEL

sdi1 0x0 xfs 0b16447b-452a-448c-826d-c2858fd1b7fa

sdi1 0xe8d7ff44000 zfs_member 7860065120403136624 disk8

sdi1 0xe8d7ff84000 zfs_member 7860065120403136624 disk8

This drive was formatted as zfs in the array a while ago, but is currently xfs. I take it this means I need to do a wipefs -a /dev/sdi1 (and to all my other drives with similar wipefs /dev/sdX1 output as above )? I've never used wipefs before so not sure of the exact procedure, consequences, if the drive needs to have all data evacuated, etc.

8 minutes ago, Vynlovanth said:

I take it this means I need to do a wipefs -a /dev/sdi1

Don't do that, that will wipe both filesystems.

If the drives are now XFS, you can wipe just the ZFS signature with

wipefs -a -t zfs_member /dev/sdi1

Note that if they are array disks and parity is valid, you should use instead the md devices with the array started in maintenance mode, or it will cause some sync errors, e.g., for disk4 it would be:

wipefs -a -t zfs_member /dev/md4p1

Aha, glad I asked.

Started my array in maintenance mode, went through my xfs formatted disks with wipefs /dev/md#p1, did a wipefs -a -t zfs_member /dev/md#p1 on any that had zfs_member mentioned in the first command, and confirmed only the xfs line remained with another wipefs /dev/md#p1.

Thank you for the assistance!