Introduction

Nodes in NotReady state with RKE2 agent service is stuck in "activating" state, waiting for agent configuration.

msg="Waiting to retrieve agent configuration; server is not ready: 
/var/lib/rancher/rke2/agent/serving-kubelet.crt: Node password rejected, duplicate 
hostname or contents of '/etc/rancher/node/password' may not match server node-passwd 
entry, try enabling a unique node name with the --with-node-id flag"

Prerequisites

• RKE2 Managed Kubernetes Cluster

Step-by-Step Instructions

1. Stop the rke2-agent service on the node that is being registered:

$ systemctl stop rke2-agent.service

2. Find the secret for the node that was not cleaned up by running the following with a kubeconfig file pointed to the RKE2 cluster in question:

$ kubectl get secret -n kube-system | grep node-password.rke2

This will output secrets that are named in the following format: <NODE_NAME>.node-password.rke2. In this article, the secret for the node in question is: ip-10-0-0-136.node-password.rke2

3. Delete the secret corresponding to the node in question:

$ kubectl delete secret -n kube-system ip-10-0-0-136.node-password.rke2

4. Start the rke2-agent service on the node that is being registered:

$ systemctl start rke2-agent.service

Additional Resources

SUSE - Unable to re-add worker node to custom RKE2 cluster