Kubeadm: Kubeadm reset should cleanup control-plane nodes properly

[fabriziopandini]

Kubeadm reset should take care of removing control-plane nodes from the ClusterStatus status ConfigMap

[Klaven]

I'll take this one.

[yagonobre]

Thanks @Klaven
/lifecycle active

[fabriziopandini]

additional info as per @Klaven request:

kubeadm-config ConfigMap is in kube-system namespace

All the machinery for reading/writing the config map already exists in cmd/kubeadm/app/phases/uploadconfig/uploadconfig.go and cmd/kubeadm/app/util/config/cluster.go

The ConfigMap has two key values: one for the ClusterConfigurationobject, one for the ClusterStatus

Whenever a control-plane node joins, a new entry is created in ClusterStatus; this PR should implement removal of entry from the Clust erStatuswhenever a kubeadm reset operation is executed on a control-plane node.

[timothysc]

/assign @Klaven

[k8s-ci-robot]

@timothysc: GitHub didn't allow me to assign the following users: Klaven.

Note that only kubernetes members and repo collaborators can be assigned.
For more information please see the contributor guide

In response to this:

/assign @Klaven

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.

[ereslibre]

Is it also desired to attempt the node deletion aside from modifying the kubeadm-config configmap?

[fabriziopandini]

@ereslibre there are multiple issue related to reset

• delete API endpoint, this issue
• delete etcd member, https://github.com/kubernetes/kubeadm/issues/1211
• delete node (discussed in slack, I'm not aware of a dedicated issue)

For sake of simplicity IMO we should treat all of them as separated issues/PR

[neolit123]

/assign