Recent Features
| |||||||||||||
ext4 and data lossext4 and data lossPosted Mar 12, 2009 18:05 UTC (Thu) by cpeterso (guest, #305)In reply to: ext4 and data loss by JoeBuck Parent article: ext4 and data loss
Why is fsync() on ext3 slower than fsync() on a non-journaled file system? Because fsync() has to update the journal on disk in addition to writing data and metadata?
(Log in to post comments)
ext3's slow fsync() Posted Mar 19, 2009 2:19 UTC (Thu) by xoddam (subscriber, #2322) [Link]
In default data=ordered mode, fsync(file) on ext3 is more-or-less equivalent to a sync of the entire filesystem. It forces all dirty file blocks to be flushed even those belonging to unrelated tasks, and the caller has to wait for the entire operation to complete. There is no clean way to flush the dirty blocks of only the desired file.
|
|||||||||||||