You need to start the balancer for redistributing data equally between all nodes so that Hadoop cluster will find new datanodes automatically. To optimize the cluster performance, you should start rebalancer to redistribute the data between datanodes.
