-
Type: Improvement
-
Resolution: Done
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: Replication, Storage
-
None
-
Storage Execution
By using the record store directly we can avoid any locking issues and use a snapshot to return data at any time.
We must be careful of operations which drop/recreate the oplog, like resync/compact/etc, since the record store pointer will have to be re-acquired once it becomes invalid.
- is related to
-
SERVER-5729 Special-case concurrency model for oplog, make reads not block as much
- Closed
-
SERVER-20328 Allow secondary reads while applying oplog entries
- Closed
- related to
-
SERVER-24661 Secondary block reader a very long time when replay oplog
- Closed