openstack_parallel_machine_learning

using terraform to provision compute cluster, and spark mlib to run parallel ML

1. run terraform to create n instances with the following characteristics:

• ssh allowed with right sec groups - port 22 open
• sec groups for 4040 and 8080 so spark web UI is available
• floating ip address associated
• large enough for required work
• running ubuntu

add all hosts to /etc/hosts scp set up scripts to hosts run set up scripts with dsh