I’m currently managing a Kubernetes cluster comprised of 3 master nodes, each running etcd and serving as a control plane. My scenario involves a theoretical situation where the entire primary data center, including all VPS instances hosting master and worker nodes, experiences a complete outage.