Atime and btrfs: a bad combination?
Atime and btrfs: a bad combination?
Posted Jun 1, 2012 8:39 UTC (Fri) by geertj (guest, #4116)Parent article: Atime and btrfs: a bad combination?
Maybe atime could be updated in-place, if atime is the only metadata update to an inode? The implication would be that atime in snapshots would change. But maybe that's something that could be accepted as a compromise?
Posted Jun 1, 2012 9:43 UTC (Fri)
by bergwolf (guest, #55931)
[Link]
Atime and btrfs: a bad combination?
Updating atime in a snapshot breaks its design and is dangerous if snapshot is taken for backup purpose.