Details
-
Task
-
Resolution: Unresolved
-
Major - P3
-
None
-
None
-
None
-
None
-
Cluster Scalability
-
34
Description
This work was done to the RSM that exists in v4.3/master, but it would be difficult to backport. I suspect the least invasive change we could make would be to use ReplicaSetMonitor::SetState::scheduleWorkAt() to explicitly notify and thus evaluate timeouts for all waiters.