Switch to the acgk8s context:
kubectl config use-context acgk8s
Check the current status of the nodes:
kubectl get nodes
Identify the broken node and save the name to a file named broken-node.txt:
echo acgk8s-worker2 > /k8s/0004/broken-node.txt
Attempt to determine the cause of the issue:
kubectl describe node acgk8s-worker2
Access the node using ssh:
ssh acgk8s-worker2
Check the kubelet log:
sudo journalctl -u kubelet
Check the last entry in the log.
Check the kubelet status:
sudo systemctl status kubelet
Enable kubelet:
sudo systemctl enable kubelet
Start kubelet:
sudo systemctl start kubelet
Verify that kubelet started successfully:
sudo systemctl status kubelet
Return to the control plane node:
exit
Check the status of the nodes:
kubectl get nodes
Check the status of the exam objectives:
./verify.sh