So if there is a spike on say errors, you do a group by pod, click on the pod in the legend, see the k8s neighborhood which means discover the pod --> config ma

So if there is a spike on say errors, you do a group by pod, click on the pod in the legend, see the k8s neighborhood which means discover the pod --> config map association. You see the config map and all the recent updates to it, and can even diff them like code. An alert on the other hand, say on a node that is low on disk, doesn't automatically tell you all the affected workloads. Not saying the alert doesn't help, but this experience helps identify the affected pods and apps much faster.