jstests/sharding/server_status_sharding_statistics.js sets a fail point to pause all find commands so it can verify statistics after migrations time out waiting for collection locks entering the config server commit phase. If the config server primary steps down while the fail point is set, mongos will automatically retry the migration which can trigger a metadata refresh that will hang because of the fail point, timing out the test. Running the test with config server stepdowns doesn't add meaningful coverage, so it should be blacklisted in the config stepdown suite to avoid this problem.