Document MachineHealthChecks for control planes

[CecileRobertMichon]

Since #3956 merged, KCP supports remediation with machine health checks. However, the documentation on the topic is missing. In fact, the current MHC docs state:

Control Plane Machines are currently not supported and will not be remediated if they are unhealthy

https://cluster-api.sigs.k8s.io/tasks/healthcheck.html#limitations-and-caveats-of-a-machinehealthcheck

We should update the existing documentation for MachineHealthChecks to include KCP.

/kind documentation

[CecileRobertMichon]

https://github.com/kubernetes-sigs/cluster-api/blob/master/docs/proposals/20191017-kubeadm-based-control-plane.md#remediation-using-delete-and-recreate can be used as a starting point

[CecileRobertMichon]

/help

[k8s-ci-robot]

@CecileRobertMichon:
This request has been marked as needing help from a contributor.

Please ensure the request meets the requirements listed here.

If this request no longer meets these requirements, the label can be removed
by commenting with the /remove-help command.

In response to this:

/help

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[CecileRobertMichon]

/milestone v0.4.0

[scottslowe]

@CecileRobertMichon I'd like to work on this. Aside from removing the text indicating that MHC doesn't apply to KCP, what other changes do you feel are necessary?

[CecileRobertMichon]

@scottslowe awesome, thank you. I think in addition to removing the text of the limitation, it would be good to add an example of how to set it up for KCP in https://cluster-api.sigs.k8s.io/tasks/healthcheck.html#creating-a-machinehealthcheck. You can use https://github.com/kubernetes-sigs/cluster-api/blob/master/test/e2e/data/infrastructure-docker/v1alpha4/cluster-template-kcp-remediation/mhc.yaml as an example.

[scottslowe]

@CecileRobertMichon In the "Limitations and Caveats" section of the current doc, there is a bullet that says "If no Node joins the cluster for a Node after the NodeStartupTimeout, the Machine will be remediated"

Should this be "If no Node joins the cluster for a Machine...", or "If no Machine joins the cluster for a Node..."? The current wording seems off.

[CecileRobertMichon]

yeah I think it should be "If no Node joins the cluster for a Machine..."

[scottslowe]

PR submitted and ready for review!