Customers have a desire to understand when OpenShift components are not performing as they normally do. This boils down to the need for some type of tool that keeps track of the baseline "performance" of a component and then some mechanism to let the component owner know when the performance envelope has been violated.

One such example is the router's CPU usage. If normal CPU usage is ~40%, but suddenly the router's CPU is at 80% for an extended period, this warrants further investigation by whoever is responsible for the router in that environment.