awaitBalancerRound works by polling on `balancerStatus` command and see when the number of rounds increase by one.
If a new primary of the configsvr is elected this approach is wrong, in fact each replicaset node of the CSRS is keeping its count of balancer rounds, so this means that there is no guarantee that the number of rounds returned by `balancerStatus` is an always increasing number.
- is related to
-
SERVER-59891 Replace the coverage from sharding_continuous_config_stepdown.yml and then delete the test suite
- Backlog