Actions
Feature #13983
openOptionally no-op flush requests when journal + cache are enabled
Status:
New
Priority:
Normal
Assignee:
-
Target version:
-
% Done:
0%
Source:
other
Tags:
Backport:
Reviewed:
Affected Versions:
Pull request ID:
Description
If both the journal and cache are enabled, it is safe to no-op flush requests since read-after-write consistency is provided by the cache and crash consistency is provided by the journal. Provide a new config option for specifying a throttled flush interval in the above case -- this determines a how many seconds of data could be lost post crash. This will help reduce the speed degradation of journaling if a user doesn't mind losing X seconds of data.
No data to display
Actions