Which is better to copy the file on worker nodes or use Sparkfiles sc.addFile? How Spark works if I copy the text file on all nodes? Do the file partition among the nodes or