Add Cluster resource template

[nprokopic]

Note: WIP, initially based on Cluster resource from cluster-aws.

What does this PR do?

(Please set a descriptive PR title. Use this space for additional explanations.)

What is the effect of this change to users?

Cluster resource is deployed.

How does it look like?

apiVersion: cluster.x-k8s.io/v1beta1
kind: Cluster
metadata:
  annotations:
    cluster.giantswarm.io/description: "Awesome Giant Swarm cluster"
  labels:
    app: "cluster"
    app.kubernetes.io/managed-by: "Helm"
    app.kubernetes.io/version: "0.0.0-dev"
    application.giantswarm.io/team: "turtles"
    giantswarm.io/cluster: "awesome"
    giantswarm.io/organization: "giantswarm"
    giantswarm.io/service-priority: highest
    cluster.x-k8s.io/cluster-name: "awesome"
    cluster.x-k8s.io/watch-filter: capi
    helm.sh/chart: "cluster-0.0.0-dev"
    cluster-apps-operator.giantswarm.io/watching: ""
  name: awesome
  namespace: default
spec:
  clusterNetwork:
    services:
      cidrBlocks:
        - 172.31.0.0/16
    pods:
      cidrBlocks:
        - 100.64.0.0/12
  controlPlaneRef:
    apiVersion: controlplane.cluster.x-k8s.io/v1beta1
    kind: KubeadmControlPlane
    name: awesome
  infrastructureRef:
    apiVersion: infrastructure.cluster.x-k8s.io/v1beta1
    kind: AWSCluster
    name: awesome

Any background context you can provide?

Working towards a PoC for restructured cluster apps (more details here).

What is needed from the reviewers?

PTAL :)

Do the docs need to be updated?

Not yet.

Should this change be mentioned in the release notes?

• CHANGELOG.md has been updated (if it exists)

[nprokopic]

It's expected that JSON schema validation is failing here, because the current validation has rules for existing cluster apps, and those rules do not work in the cluster chart.

[njuettner]

I know this PR is only taking the source from https://github.com/giantswarm/cluster-aws/blob/master/helm/cluster-aws/templates/_cluster.tpl. Feel free to merge 👍🏻

[nprokopic]

We would statically set these from the cluster-$provider app in its Helm values like this:

cluster:
  internal:
    ...

and these would be treated as an internal API, as in internal in our apps, where the cluster chart exposes it to the cluster-$provider , but it's not meant to be set by end users. This is roughly aligned with the current behavior of Values.internal.

[nprokopic]

I know this PR is only taking the source from...

In a nutshell all initial PRs will be like that, with "just" make the templates being nicely configurable, so all providers can use them, so we have more work around Helm values and JSON schema in order to make this work nicely.