@geerlingguy you should think about adding to cluster - a scalable cluster - imagine how many you could fit into the tiny rack - 2 discrete clusters each with its own ups - they could probably do some real work and be reliable - you could always sin up some vms to join cluster if you have workload spikes