The most common source of metric duplication is because of the combination of:
- The Cluster Metrics feature
- ServiceMonitors already existing that scrape an existing kube-state-metrics, kubelet, cadvisor, node exporter.
We should try to detect a few of these cases and warn in the NOTES.tpl (not an error, but a warning).