Allow the connector to run outside of a kube pod

[dnephin]

Summary

This is still pretty rough, but with these changes I was able to run a connector locally, and have it connector to a server that I'm also running locally. I was getting some strange errors about error: the server doesn't have a resource type "pods" when trying to use the connector to connect to kube, but that might be my setup. I'm guessing that's a 404 being translated into a "helpful" error message.

Resolves #2500

[dnephin]

but keeping support for Kubernetes while in this mode seems unnecessary

The goal here isn't to support different kinds of connectors, but to allow the kubernetes connector to run in different environments. For "getting started" and while developing/contributing, being able to run the connector locally outside of a kube pod is a big improvement. It removes all the complications of helm, and makes it much easier to iterate and try things out.

Even in production environments, having the authentication proxy run inside the environment for which it is authenticating requests could cause problems, especially when trying to recover that environment.

Perhaps what is needed is to decouple the connector from the Kubernetes connector so it can be run separately with possibly other integrations

I think that would also be a good change, but there's not much value in doing that until we are ready to add a different kind of destination.