Hey folks,

 

My cache consists of 2 1TB SSDs in a RAID1 BTRFS pool. The main GUI shows 0 errors on both drives, however my log is telling a different story...

 

Nov 21 21:19:32 Gandalf kernel: BTRFS critical (device sdb1): corrupt leaf: root=5 block=6601968369664 slot=126 ino=17077501 file_offset=24920064, invalid num_bytes for file extent, have 281472964518981, should be aligned to 4096
Nov 21 21:19:36 Gandalf kernel: BTRFS critical (device sdb1): corrupt leaf: root=5 block=6601968369664 slot=126 ino=17077501 file_offset=24920064, invalid num_bytes for file extent, have 281472964518981, should be aligned to 4096
Nov 21 21:19:40 Gandalf kernel: BTRFS critical (device sdb1): corrupt leaf: root=5 block=6601968369664 slot=126 ino=17077501 file_offset=24920064, invalid num_bytes for file extent, have 281472964518981, should be aligned to 4096
Nov 21 21:19:44 Gandalf kernel: BTRFS critical (device sdb1): corrupt leaf: root=5 block=6601968369664 slot=126 ino=17077501 file_offset=24920064, invalid num_bytes for file extent, have 281472964518981, should be aligned to 4096
Nov 21 21:19:48 Gandalf kernel: BTRFS critical (device sdb1): corrupt leaf: root=5 block=6601968369664 slot=126 ino=17077501 file_offset=24920064, invalid num_bytes for file extent, have 281472964518981, should be aligned to 4096
Nov 21 21:19:52 Gandalf kernel: BTRFS critical (device sdb1): corrupt leaf: root=5 block=6601968369664 slot=126 ino=17077501 file_offset=24920064, invalid num_bytes for file extent, have 281472964518981, should be aligned to 4096

This error repeats constantly.

 

I ran a scrub on the cache which shows 4 uncorrectable errors. Strangely enough the main GUI still shows 0 errors and I've not noticed any other symptoms.

 

How do I go about fixing this? 

 

gandalf-diagnostics-20181121-2116.zip

I'm not by any means a BTRFS expert so I'll refer you to a post written by a guy who is. I believe you're going to have to backup as much data as you can, reformat and then restore it.