A btrfs filesystem reported some errors during btrfs receive operation due to some missing extends. From the log:
BTRFS error (device dm-1): unable to find ref byte nr 190303420416 \
parent 0 root 594 owner 1 offset 0
BTRFS: error (device dm-1) in __btrfs_free_extent:6944: errno=-2 No such entry
BTRFS info (device dm-1): forced readonly
BTRFS: error (device dm-1) in btrfs_run_delayed_refs:2956: errno=-2 No such entry
A btrfs scrub (after a umount + mount) failed with similar errors.
A btrfs check reports issues for 3 extends - each extend issue is reported like this:
checking extents
ref mismatch on [190303420416 16384] extent item 0, found 1
Backref 190303420416 parent 594 root 594 not found in extent tree
backpointer mismatch on [190303420416 16384]
owner ref check failed [190303420416 16384]
My question is: how can I translate those numbers to something useful? For example to check if some files/directories are impacted?
And how to deal with such errors?
A btrfs check --repair seems to work without complaining severely:
ref mismatch on [190303420416 16384] extent item 0, found 1
* repair deleting extent record: key 190303420416 169 1
* adding new tree backref on start 190303420416 len 16384 parent 0 root 594
Backref 190303420416 parent 594 root 594 not found in extent tree
backpointer mismatch on [190303420416 16384]
owner ref check failed [190303420416 16384]
(the * marks are mine)
Does this mean that the repair was successful without loosing any data?