Add a Prometheus metric that measures current cache memory utilization

[timflannagan]

Introduce a custom Prometheus gauge metric that increase visibility into the core rukpak provisioners memory usage. We're able to deduce the current memory footprint through overall usage, and dive into the provisioner's container to determine how much the /var/bundles/cache directory is currently using, but less visibility into how large the dynamic GVK/watch cache size is during steady state events.

It would be nice to have a cache size metric that covers both the bundle cache size and the dynamic cache sizes.

[asmacdo]

@timflannagan recommends backlog

[github-actions[bot]]

This issue has become stale because it has been open 60 days with no activity. The maintainers of this repo will remove this label during issue triage or it will be removed automatically after an update. Adding the lifecycle/frozen label will cause this issue to ignore lifecycle events.