https://bugzilla.suse.com/show_bug.cgi?id=1222346
https://bugzilla.suse.com/show_bug.cgi?id=1222346#c6
--- Comment #6 from Wenruo Qu ---
I'm wondering if snapper is not handling rescan correctly.
Yes, qgroup is very heavy, but that's so far the only way to get accurate
accounting numbers, considering how many tricks we put to get super fast
snapshotting.
All the super fast snapshotting has put a lot of complexity to qgroups.
Not to mention the extra complexity from multi-level qgroups, utilized by
snapper.
But for the RESCAN_WAIT returns immeidately problem, I think it's something
from the snapper side.
Snapper should check if qgroup is even enabled and rescan is running.
Or RESCAN_WAIT ioctl immediately return 0 if qgroup rescan is not running nor
qgroup is enabled.
Any clue how snapper triggers qgroup rescan? Did it check if qgroup is enabled
or handles qgroup rescan return value correctly?
--
You are receiving this mail because:
You are on the CC list for the bug.