I have two 250 GB NMVE drives setup as my cache drive and it looks like a checksum error is forcing one of them (effectively both) into read-only mode.

 

Can anyone help with how to fix this issue?

 

Dec 23 06:44:01 dvl-lotf kernel: BTRFS warning (device nvme0n1p1): nvme0n1p1 checksum verify failed on 1278820352 wanted 0x4e4e974a found 0xe7174b2f level 0
Dec 23 06:44:01 dvl-lotf kernel: BTRFS: error (device nvme0n1p1) in add_to_free_space_tree:1039: errno=-5 IO failure
Dec 23 06:44:01 dvl-lotf kernel: BTRFS info (device nvme0n1p1): forced readonly
Dec 23 06:44:01 dvl-lotf kernel: BTRFS: error (device nvme0n1p1) in __btrfs_free_extent:3221: errno=-5 IO failure
Dec 23 06:44:01 dvl-lotf kernel: BTRFS: error (device nvme0n1p1) in btrfs_run_delayed_refs:2144: errno=-5 IO failure

 

Diagnostics attached.

lotf-diagnostics-20211223-0743.zip

Best bet is to backup and reformat the pool, you can also try btrfs fsck, but only after everything is backed up.

Should I just backup everything in /mnt/cache or is there another directory where the files would be?

38 minutes ago, mwcarroll said:

/mnt/cache

That's it.

Thanks! This solved my issue.