You are troubleshooting an issue in your Kubernetes cluster where worker nodes are intermittently unable to communicate with the control plane. You suspect an issue with the Kubernetes API server, as some components of the control plane are becoming unreachable. The cluster consists of three master nodes in an HA configuration, and each node is configured with an etcd database. Which of the following could most likely explain the intermittent communication issues between the worker nodes and the control plane?