Using vector time to handle multiple failures in distributed systems | IEEE Journals & Magazine | IEEE Xplore