This is two http applications designed to prove whether networking is possible between kubernetes worker nodes. We made it because at one point after an upgrade the network appeared to be not always working, but we didn't really have a good way to run diagnostics, and we never figured out exactly what was going on.
With this tool, we'll be able to have 100% confidence that containers on one node are able (or are not able) to talk to containers on other nodes. Here's a diagram:
Prometheus is itself out of scope; the monitors should be able to export metrics that prometheus understands; and the indicators should be accessible by the monitors so that the monitors can verify working network connectivity.