The prometheus endpoint should expose more metrics that allow to detect whether stackset controller is running well. Some possibilities: * number of errors / error rate * number of stacksets / stacks under control