mikejoh / k8s-on-kvm

Deploy small clusters for learning purposes e.g. CKA studies

Kubernetes on KVM

Pre-requisites

• KVM
• OpenTofu (the terraform fork)

Notes

• See the scripts/ folder for various utility scripts.
• You probably want to change user and group to libvirt-qemu and kvm respectively in /etc/libvirt/qemu.conf to mitigate permission issues on storage pools.

Getting started

Provision cluster nodes

By default we'll deploy a cluster on three nodes, they will have both the control-plane and worker roles.

If you're setting a lower resource values on each node then you might need to set: --ignore-preflight-errors=mem,numcpu during kubeadm init.

1. Change the k8s.auto.tfvars to fit your needs!
2. Run tofu init
3. Run tofu plan
4. Run tofu apply
5. SSH to all nodes using the private and public key pair you referenced when deploying the cluster. You can find the IP addresses of the cluster nodes by running:

sudo virsh net-dhcp-leases k8s_net

6. Proceed with the bootstrapping the Kubernetes cluster using e.g. kubeadm.

Install a CNI plugin:

Flannel CNI:

kubectl apply -f https://github.com/flannel-io/flannel/releases/latest/download/kube-flannel.yml

Calico CNI:

kubectl apply -f https://raw.githubusercontent.com/projectcalico/calico/v3.25.0/manifests/calico.yaml

Cilium CNI: https://docs.cilium.io/en/stable/installation/k8s-install-kubeadm/#deploy-cilium

Create a multi-node control plane (for high availability)

Add control-plane nodes to the cluster:

echo "$(kubeadm token create --print-join-command) --control-plane --certificate-key $(kubeadm init phase upload-certs --upload-certs --skip-headers --skip-log-headers 2>/dev/null | tail -n 1)"

Add worker nodes to the cluster:

1. Generate the join command on a control-plane node:

kubeadm token create --print-join-command

2. Use the generate join command and run that on the worker node.

Upgrade a cluster

On the first control-plane node:

1. Upgrade kubeadm:

export NEXT_VERSION="1.26.2"

apt-mark unhold kubeadm
apt-get update
apt-get install -y kubeadm=${NEXT_VERSION}-00
apt-mark hold kubeadm

2. Check the upgrade plan:

kubeadm upgrade plan

3. Apply the upgrade plan:

kubeadm upgrade apply ${NEXT_VERSION}

4. Drain the node:

kubectl drain <node-to-drain> --ignore-daemonsets

5. Upgrade kubectl and kubelet:

apt-mark unhold kubelet kubectl
apt-get update
apt-get install -y kubelet=${NEXT_VERSION}-00 kubectl=${NEXT_VERSION}-00
apt-mark hold kubelet kubectl

6. Restart the services:

sudo systemctl daemon-reload
sudo systemctl restart kubelet

7. Uncordon the node to allow scheduling again:

kubectl uncordon <node-to-uncordon>

8. Repeat on the rest of the control-plane nodes!

Clean up the cluster

Run the clean-up utility script: scripts/clean_up.sh

k8s-on-kvm's Issues

feature: Use terraform with ansible provider to make changes needed to VMs?