There's a bunch of stuff running on the cluster; I need to shut down the scheduler and will be shutting down the nodes before too long. we may have to restart the things that are currently running; if we're lucky the scheduler will restart them for us. worst case we just restart the running jobs. --bob