Scale out nodes

Currently there is no redundancy available in nodes such that if one node fails all applications can be run on another. I should build out an additional slave node so that applications can fail across nodes and there is space to run all applications twice.