I have a running job on the Mesos Spark cluster. I would like to add a new slave to the cluster without stopping the master, and allocate the new added slave to the running job.
If you bring up a slave with the --master flag set to the existing master it should automatically get added to the cluster. Any tasks that are already running on the slave will not get rescheduled on this new slave, but any tasks launched by the spark job will use the resources provided by this new slave.