Display pod counts/limits on Cluster monitoring page
We should display a few key metrics regarding pods on the Cluster monitoring page:
- Pods per Node (Average): The average number of pods per node.
- Pods per Node (Peak): The highest number of pods on any node.
- Pods per Node (Capacity): The total number of allowed pods per node. Usually 100.
The intent here is to give cluster admins an idea if they are running out of pod space. This can happen on clusters with workloads of many small pods. It's also possible to have a specific node type that is in high demand (say GPU nodes) which may cap out much more rapidly than other types. So it's helpful to have both the average and peak amount.