We want to add an additional Telemetry metric that may be used for consumption based billing. This metric will be very similar to the existing pod runtime metric, but multiplied by the number of CPU limits for each notebook pod.

Example:

If I notebook pod had a CPU limits value of 2 and runs for 4 hours, that should yield a value of 8 for this metric (2*4)

We should be able to use a calculation like the following to get the number of CPU requests for the pod:

kube_pod_container_resource_limits{namespace='rhods-notebooks', resource='cpu'}