As part of our metrics story it might be useful to also get ctdb level metrics, similar to issue #46 but specific to things like ctdb health, etc.