This issue may happen if the
1) config shard is being shut down - the secondaries are successfully shut down, the primary Balancer is in the main loop and waits for ShardRegistry::reload

2) ShardRegistry::reload is already dispatched to the internalReload

3) As its a config shard the reload is using LocalShard to do exhaustiveFindOnConfig

4) the exhaustiveFindOnConfig calls for the ReplicationCoordinator::waitUntilOpTimeForRead

5) the last call never returns as the secondaries have stepped down already