A while ago we implemented a ssa.Cache to cache SSA calls. I think it would be great if we have corresponding metrics (e.g. cache hit/miss) so we can observe how well the cache is working.

We should at least have labels for:

• the object of the SSA call (probably just kind)
• the controller in which the cache is used (MD / MS / cluster topology / ...)

I think we'll have to extend the cache a bit to have the data for the labels.

Also let's take a look at how cache metrics usually look like in the Prometheus format and try to be consistent with that.