I have a query regarding the RKE2 cluster control and worker nodes replacement (drain the node , delete the node and then the new node that comes up joins to the cluster) . I want these operations to be part of an operator pattern. I want to know if rancher have already controllers written in the rancher codebase. I saw - https://github.com/rancher/rancher/blob/main/pkg/controllers/management/node/cleanup.go where there is a controller that does have a function of drain Node. Can someone help me point to any documentation to enable these controllers ? I want the cycling of control/worker nodes within cluster itself via operator pattern of Kubernetes. Also this is the blog ( https://www.databricks.com/blog/scalable-kubernetes-upgrade-using-operators ) that is helpful in the same context but I need some more guidance to actually implement it or if there is already an implementation - please tell the same.