Upgrading Kubernetes Cluster
============================
Start the upgrade by performing Step 1 to 13 on one of the master nodes (if more than 1 master nodes). Step 14 is to be performed on all other nodes (the remainder master and worker nodes).

1. Update the repository version reference
sed -i 's/33/34/g' /etc/apt/sources.list.d/kubernetes.list

apt update

2. View available versions of kubeadm
apt-cache madison kubeadm

3. Upgrade kubeadm
apt-mark unhold kubeadm
apt install -y kubeadm=1.34.1-*
apt-mark hold kubeadm

4. Verify kubeadm version
kubeadm version

5. Evict pods from node. Daemonsets such as cilium need to be on every node.
kubectl drain master --ignore-daemonsets

6. Plan the upgrade. Read through the output
kubeadm upgrade plan

7. Apply the upgrade. Answer y when prompted.
kubeadm upgrade apply v1.34.1

8. Check the status of nodes. Why is the version still the same? Why is SchedulingDisabled?
kubectl get node

9. Upgrade kubelet and kubectl
apt-mark unhold kubelet kubectl
apt install -y kubelet=1.34.1-* kubectl=1.34.1-*
apt-mark hold kubelet kubectl

10. Restart the daemon
systemctl daemon-reload
systemctl restart kubelet

11. Check status. Notice that worker is still on the old version
kubectl get nodes

12. Re-enable scheduling for master node
kubectl uncordon master

13. Confirm master node status. It should be Ready.
kubectl get nodes

14. Repeat steps 1 to 6, for all nodes (including other control plane nodes), followed by
    "kubeadm upgrade node", then 9 to 14. solution as below:

sed -i 's/33/34/g' /etc/apt/sources.list.d/kubernetes.list
apt update
apt-cache madison kubeadm
apt-mark unhold kubeadm
apt install -y kubeadm=1.34.1-*
apt-mark hold kubeadm
kubeadm version
kubectl drain worker --ignore-daemonsets
kubeadm upgrade node
kubectl get node
apt-mark unhold kubelet kubectl
apt install -y kubelet=1.34.1-* kubectl=1.34.1-*
apt-mark hold kubelet kubectl
systemctl daemon-reload
systemctl restart kubelet
kubectl get nodes
kubectl uncordon worker
kubectl get nodes