We have Spark 1.0.1 on Mesos deployed as a cluster in EC2. Our Devops lead tells me that Spark jobs can not be submitted from local machines due to the complexity of opening the right ports to the world etc.
Are other people running the shell locally in a production environment?
