Problem with BTRFS filesystem

Greetings all,

recently I noticed my rsync backups were having errors, so I've investigated and here is what I'v found until now:

from status.log it seems the problem occured just after the defragmentation on 12-09-19:

[19/12/07 23:35:44 CET] notice:backup:LOGMSG_SUCCESS_BACKUP Successfully completed backup job Backup Readynas-Pro-6.
[19/12/08 01:00:09 CET] notice:system:LOGMSG_SYSTEM_HALT The system is shutting down.
[19/12/08 08:01:47 CET] info:system:LOGMSG_START_READYNASD ReadyNASOS background service started.
[19/12/09 00:03:00 CET] err:backup:LOGMSG_ERROR_BACKUP Error backup job Backup Readynas-Pro-6: Failure during copy.
[19/12/09 01:00:09 CET] notice:system:LOGMSG_SYSTEM_HALT The system is shutting down.
[19/12/09 01:00:09 CET] warning:system:LOGMSG_SENT_ALERT_MESG_FAILED Alert message failed to send.
[19/12/09 07:01:45 CET] info:system:LOGMSG_START_READYNASD ReadyNASOS background service started.
[19/12/09 08:00:01 CET] notice:volume:LOGMSG_DEFRAGSTART_VOLUME Defragmentation started for volume data.
[19/12/09 08:00:01 CET] notice:volume:LOGMSG_DEFRAGSTART_VOLUME Defragmentation started for volume data.
[19/12/09 08:02:50 CET] notice:volume:LOGMSG_DEFRAGEND_VOLUME Defragmentation complete for volume data.
[19/12/10 01:00:02 CET] notice:system:LOGMSG_SCHEDULED_POWEROFF_DELAY Scheduled power off was delayed.
[19/12/10 01:00:28 CET] err:backup:LOGMSG_ERROR_BACKUP Error backup job Backup Readynas-Pro-6: Failure during copy.
[19/12/10 01:03:10 CET] notice:system:LOGMSG_SYSTEM_HALT The system is shutting down.
[19/12/10 07:01:42 CET] info:system:LOGMSG_START_READYNASD ReadyNASOS background service started.

on 12-06-19 I executed a "balance" on the volume without any problem, here is the volume.log

data        balance    2019-12-06 13:05:36  2019-12-06 13:06:37  completed  Done, had to relocate 21 out of 5634 chunks 

but now BTRFS filesystem seems to be corrupted, kernel.log:

Dec 22 23:38:51 Readynas-Pro-6 kernel: BTRFS critical (device md125): corrupt leaf: root=1 block=5563987623936 slot=63 ino=4814786, invalid dir item type for XATTR key, have 1 expect 8
Dec 22 23:38:51 Readynas-Pro-6 kernel: BTRFS error (device md125): error loading props for ino 4814775 (root 275): -5

systemd-journal.log

Dec 25 00:10:00 Readynas-Pro-6 kernel: BTRFS critical (device md125): corrupt leaf: root=1 block=5563987623936 slot=63 ino=4814786, invalid dir item type for XATTR key, have 1 expect 8

kernel.log

Dec 22 23:38:51 Readynas-Pro-6 kernel: BTRFS critical (device md125): corrupt leaf: root=1 block=5563987623936 slot=63 ino=4814786, invalid dir item type for XATTR key, have 1 expect 8
Dec 22 23:38:51 Readynas-Pro-6 kernel: BTRFS error (device md125): error loading props for ino 4814775 (root 275): -5

So now each rsync backup is terminated with errors ! That's not 100% critical for now, but I thought this filesystem was far more robust !

backup_001_copy.log

rsync: readlink_stat("/tmp/12264/Stockage-Marc/dummy.example1") failed: Input/output error (5)
rsync: readlink_stat("/tmp/12264/Stockage-Marc/dummy.example2") failed: Input/output error (5)
IO error encountered -- skipping file deletion

My ReadyNAS-Pro-6 have been working like a charm with OS6 for several years without any serious problem, it's on UPS and was not forced to shutdown at this time. It already happened I had to force a reset, but at worst it resynchronized and that's all.

Scanning the disks didn't solve the problem at all.

So if anybody have a suggestion (excepted a factory reset), I'd be glad to hear it.

Thanks in advance !