What you just said is fun, because that is a design choice in Kubernetes. You DO have a virtual hierarchical configuration. It’s just that it’s not a txt file,

What you just said is fun, because that is a design choice in Kubernetes. You DO have a virtual hierarchical configuration. It’s just that it’s not a txt file, it’s a json object. The whole Kubernetes things is actually that: a single json defining a supercomputer. Changes to the json become changes in the supercomputer. The whole complexity aims just that: make that task easy. You have a lot of tools to interact with that json, but you actually only need a http client, because it is exposed as an API.