[cpretzer]

This just shows the importance of day 2 operations. The first step is simply having the ability to collect the metrics, making those metrics actionable is the "leveling up" that it takes to run an application on Kubernetes.

Also, big +1 for the prometheus queries in this post. I still am trying to get my head around PromQL

[GauntletWizard]

I cannot stress this enough. Gather the right metrics, and you can analyze them how and when you wish. Gather the wrong ones, and there’s no coming back.