This is a low priority but having the ability to run multiple masters that
synch important information so the cluster stays up if one fails. We could also
delegate jobs/tasks and the collection of output between them if we need to
handel larger jobs.
Original issue reported on code.google.com by ryanbres...@gmail.com on 27 Jul 2012 at 11:02
Original issue reported on code.google.com by
ryanbres...@gmail.comon 27 Jul 2012 at 11:02