DEV Community

Cover image for Upgrading a Kubernetes v1.28 Cluster using kubeadm
Unni P
Unni P

Posted on • Updated on • Originally published at iamunnip.hashnode.dev

Upgrading a Kubernetes v1.28 Cluster using kubeadm

#kubernetes #cloudnative #opensource #beginners

In this article, we will look how we can upgrade a three node Kubernetes v1.28 cluster using kubeadm

Prerequisites

  • We need a working Kubernetes cluster created using kubeadm

  • If you haven’t created the cluster, please check my previous article below
    This will guide you to create a Kubernetes v1.28.0 cluster
    Building a Kubernetes v1.28 Cluster using kubeadm

Upgrade Process

Control Plane

Verify the version of our cluster

# control-plane

kubectl get nodes
NAME            STATUS   ROLES           AGE    VERSION
control-plane   Ready    control-plane   3m6s   v1.28.0
node-1          Ready    <none>          43s    v1.28.0
node-2          Ready    <none>          37s    v1.28.0
Enter fullscreen mode Exit fullscreen mode

Update the package index and find the latest kubeadm patch available in the repository

# control-plane

sudo apt update

sudo apt-cache madison kubeadm
Enter fullscreen mode Exit fullscreen mode

Unhold the kubeadm package for upgrading

# control-plane

sudo apt-mark unhold kubeadm
Enter fullscreen mode Exit fullscreen mode

Upgrade the kubeadm package to version v1.28.1 and hold the package from automatic upgrading

# control-plane

sudo apt install -y kubeadm=1.28.1-1.1

sudo apt-mark hold kubeadm
Enter fullscreen mode Exit fullscreen mode

Check the upgraded kubeadm version and verify the upgrade plan

# control-plane

sudo kubeadm version

sudo kubeadm upgrade plan
Enter fullscreen mode Exit fullscreen mode

Once our plan is verified, we can upgrade the control-plane by executing the below command

# control-plane

sudo kubeadm upgrade apply v1.28.1
Enter fullscreen mode Exit fullscreen mode

Prepare the control-plane for maintenance by marking it unschedulable and evicting the workloads

# control-plane

kubectl drain control-plane --ignore-daemonsets
Enter fullscreen mode Exit fullscreen mode

Unhold the kubelet and kubectl packages for an upgrade

# control-plane

sudo apt-mark unhold kubelet kubectl
Enter fullscreen mode Exit fullscreen mode

Upgrade kubelet and kubectl packages to version v1.28.1 and hold the packages from automatic upgrading

# control-plane

sudo apt install kubelet=1.28.1-1.1 kubectl=1.28.1-1.1

sudo apt-mark hold kubelet kubectl
Enter fullscreen mode Exit fullscreen mode

Restart the kubelet service on control-plane

# control-plane

sudo systemctl daemon-reload

sudo systemctl restart kubelet
Enter fullscreen mode Exit fullscreen mode

Uncordon the control-plane for marking it as schedulable

# control-plane

kubectl uncordon control-plane
Enter fullscreen mode Exit fullscreen mode

Verify the nodes, now we can see control-plane is upgraded to v1.28.1

# control-plane

kubectl get nodes
NAME            STATUS   ROLES           AGE     VERSION
control-plane   Ready    control-plane   9m17s   v1.28.1
node-1          Ready    <none>          6m54s   v1.28.0
node-2          Ready    <none>          6m48s   v1.28.0
Enter fullscreen mode Exit fullscreen mode

Nodes

Update the package index on node-1 and node-2 and unhold the kubeadm package for upgrading

# node-1 and node-2

sudo apt update

sudo apt-mark unhold kubeadm
Enter fullscreen mode Exit fullscreen mode

Upgrade the kubeadm package to version v1.28.1 on node-1 and node-2 and hold the package from automatic upgrading

# node-1 and node-2

sudo apt install kubeadm=1.28.1-1.1

sudo apt-mark hold kubeadm
Enter fullscreen mode Exit fullscreen mode

Upgrade the cluster configuration on node-1 and node-2 using the below command

# node-1 and node-2

sudo kubeadm upgrade node
Enter fullscreen mode Exit fullscreen mode

Prepare node-1 and node-2 for maintenance by marking them as unschedulable and evicting the workloads

Execute the below drain command on control-plane

# control-plane

kubectl drain node-1 --ignore-daemonsets --force

kubectl drain node-2 --ignore-daemonsets --force
Enter fullscreen mode Exit fullscreen mode

Unhold the kubelet and kubectl packages for an upgrade

# node-1 and node-2

sudo apt-mark unhold kubelet kubectl
Enter fullscreen mode Exit fullscreen mode

Upgrade kubelet and kubectl packages to version v1.28.1 and hold the packages from automatic upgrading

# node-1 and node-2

sudo apt install kubelet=1.28.1-1.1 kubectl=1.28.1-1.1

sudo apt-mark hold kubelet kubectl
Enter fullscreen mode Exit fullscreen mode

Restart the kubelet service on node-1 and node-2

# node-1 and node-2

sudo systemctl daemon-reload

sudo systemctl restart kubelet
Enter fullscreen mode Exit fullscreen mode

Uncordon node-1 and node-2 for marking it as schedulable

Execute the below uncordon command on the control-plane

# control-plane

kubectl uncordon node-1

kubectl uncordon node-2
Enter fullscreen mode Exit fullscreen mode

Now our cluster is fully upgraded to v1.28.1

Verify the same by executing the below command on the control-plane

# control-plane

kubectl get nodes
NAME            STATUS   ROLES           AGE   VERSION
control-plane   Ready    control-plane   14m   v1.28.1
node-1          Ready    <none>          11m   v1.28.1
node-2          Ready    <none>          11m   v1.28.1
Enter fullscreen mode Exit fullscreen mode

Reference

https://kubernetes.io/docs/tasks/administer-cluster/kubeadm/kubeadm-upgrade/

Top comments (0)

Read next

evotik profile image

8 Essential VS Code Extensions [2024]

Evotik -

gaspardb profile image

What we learned building our SaaS with Rust 🦀

Gaspard Boursin -

yuricodesbot profile image

Supabase Security Advisor & Performance Advisor

Yuri -

kueiapp profile image

Start to use VueJS

Kueiapp -