I realized today when explaining the Cloud Native principle about externalizing control and observability in order to enable automation that it basically boils down to https://en.wikipedia.org/wiki/Fundamental_theorem_of_software_engineering …: it can be solved by another layer of indirection
-
Are you suggesting running Prometheus inside the Kubernetes cluster is not a good idea?
No, I was referring to a general design principle mentioned in https://github.com/cncf/toc/blob/master/DEFINITION.md …. VMs->containers, raw networking -> service mesh, business logic -> functions, imperative -> declarative all enable intelligent orchestration
