Hello everyone,

 

Sorry to bother, 

I have a drive on my 3 datadisk which doesn't allow writing on it anymore.

I've looked on many page of this forum prior to writing this, but I don't know what to do anymore.

 

When I start my array this shows up: 

Mar  4 14:15:18 Tower kernel: BTRFS info (device loop2): enabling ssd optimizations
Mar  4 14:15:18 Tower kernel: BTRFS info (device loop2): auto enabling async discard
Mar  4 14:15:18 Tower root: Resize device id 1 (/dev/loop2) from 20.00GiB to max
Mar  4 14:15:19 Tower emhttpd: shcmd (520): /etc/rc.d/rc.docker start
Mar  4 14:15:19 Tower rc.docker: Starting Docker daemon...
Mar  4 14:15:27 Tower kernel: BTRFS info (device md3p1): found 27 extents, stage: move data extents
Mar  4 14:15:28 Tower kernel: BTRFS critical (device md3p1): unable to find chunk map for logical 9836243806467424256 length 4096
Mar  4 14:15:28 Tower kernel: BTRFS critical (device md3p1): unable to find chunk map for logical 9836243806467424256 length 16384
Mar  4 14:15:28 Tower kernel: BTRFS critical (device md3p1): unable to find chunk map for logical 9836243806467424256 length 16384
Mar  4 14:15:28 Tower kernel: BTRFS error (device md3p1): failed to run delayed ref for logical 174850048 num_bytes 16384 type 182 action 1 ref_mod 1: -5
Mar  4 14:15:28 Tower kernel: BTRFS error (device md3p1: state A): Transaction aborted (error -5)
Mar  4 14:15:28 Tower kernel: BTRFS: error (device md3p1: state A) in btrfs_run_delayed_refs:2184: errno=-5 IO failure
Mar  4 14:15:28 Tower kernel: BTRFS info (device md3p1: state EA): forced readonly
Mar  4 14:15:28 Tower kernel: BTRFS info (device md3p1: state EA): balance: ended with status: -30
Mar  4 14:16:14 Tower rc.docker: Docker daemon...  Started.

 

Attached is my anonymized diagnostic.

And just to addup: I can still write files on my cache, the mover won't move to the "bad" disk anymore.

It's just on the 3rd disk that i can't.

 

What should i do ?

 

Thanks for your time.

tower-diagnostics-20250304.zip

There is filesystem corruption on disk3, and with btrfs I would recommend backing up the disk, reformatting and restoring the data.

 

You could also try repairing the filesystem, but I would only do that after a backup is done, since it can make things worse.

Hello,

Thanks for the quick answer, i was hoping it was not the case :'(.

 

At the time i put up my Unraid it was said that BTRFS for disk drives were the best choice, is it still the case ? or should i opt out for something else ?

If the sole solution is to format i would prefer to put the better one.

 

Is there a functionality to backup the FS to another disk ? by changing the parameter of the shares from disk3 to disk2 and apply the mover for example ? or something simpler i didn't think of ?

 

Best regards

6 minutes ago, Skold said:

At the time i put up my Unraid it was said that BTRFS for disk drives were the best choice,

We recommend XFS for the array for the typical users, unless you care about some of the btrfs extra features, like checksum or snapshots.

 

Mover cannot delete form the source since it's read only, I would recommend just copying using your favorite tool.