Improved automatic monitoring

[iangray001]

The current state of the cluster is currently not visible for monitoring.

The containers could report into a suitable monitoring service to ensure uptime of the relay and all of the boards that are mentioned in vlab.conf.

Equally, just because a boardserver is alive doesn't mean the FPGA is. When not in use, a service should periodically connect to each board, attempt to program the FPGA, and (if the board is a Zynq) attempt to send a binary, in order to check that everything is working as expected.