-
Type: Task
-
Resolution: Done
-
Priority: Major - P3
-
Affects Version/s: None
-
Component/s: Concurrency, Replication
-
None
-
Minor Change
Currently shutdown starts by waiting for at least 1 secondary to be within 10 seconds of its optime (while still accepting new writes), then it steps down, then it waits again for a secondary to be completely caught up before shutting down. Once the new concurrency system allows it, we should start by blocking new incoming writes while retaining PRIMARY status, and wait until a secondary is completely caught up before stepping down and then shutting down.
- is related to
-
SERVER-15858 Don't block writes in stepDown unless you believe there is at least 1 node within 10 seconds of your optime
- Closed
- related to
-
SERVER-14444 Implement stepdown functionality in the new ReplicationCoordinator
- Closed
-
SERVER-15861 Add argument to replSetStepDown to allow users to specify how long to wait for secondaries to catch up
- Closed