Details
-
Bug
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
Description
For volume/bucket table currently it is full cache, and we need to cleanup entries only when they are marked for delete.
So, it is unnecessary to call cleanup and waste the CPU resources on OM.
Similarly for other tables, when the double buffer has transaction entries that touch those tables, then only call cleanup.
Attachments
Issue Links
- relates to
-
HDDS-2309 Optimise OzoneManagerDoubleBuffer::flushTransactions to flush in batches
- Open
- links to