Setup a Kubernetes learning environment, optimized in Mainland China (You know it :>).

This project helps you setup your own Kubernetes cluster, current running version is v1.20.

Prerequisite

• Virtuabox 6.1.26
• Vagrant 2.2.18
• Packer 1.7.6
• A host with 24+ GiB memory (for 3 hosts setup)

Build your own box

1. Execute the command to build the box.

packer build .

2. Import to your vagrant box list.

vagrant box add --name basek8s box/package.box

Use the box created above

1. Create a new host network in Virtualbox.

From menu item File -> Host Network Manager, create a new adapter named vboxnet1, assign IPv4 address 192.168.34.1 and network mask 255.255.255.0 to it.

This network will be used by cluster nodes.

2. Start the VMs.

vagrant up

When VMs are up, initialize the Kubernetes cluster via kubeadm, and make sure it has the correct node IP set.

vagrant ssh primary

# make sure the network is working.
ping -c 2 192.168.34.3
ping -c 2 192.168.34.4

sudo kubeadm init --config kubeadm-init.yaml

# copy KUBECONFIG to $HOME/.kube/config as the output of previous step noted.

kubectl get no -o wide

# make sure the node internal-ip is 192.168.34.2 here,
# but the node status is NotReady.

NAME      STATUS     ROLES                  AGE   VERSION    INTERNAL-IP    EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION     CONTAINER-RUNTIME
primary   NotReady   control-plane,master   75s   v1.20.11   192.168.34.2   <none>        Ubuntu 20.04.3 LTS   5.4.0-88-generic   docker://20.10.9

3. Install a CNI plugin, here we choose cilium.

helm repo add cilium https://helm.cilium.io/
helm install cilium cilium/cilium --version 1.10.4 --namespace kube-system

# wait for a while depending on your network, when it's successfully setup, the primary node should become to Ready.

kubectl get no -o wide

NAME      STATUS     ROLES                  AGE     VERSION    INTERNAL-IP    EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION     CONTAINER-RUNTIME
primary   Ready   control-plane,master   5m29s   v1.20.11   192.168.34.2   <none>        Ubuntu 20.04.3 LTS   5.4.0-88-generic   docker://20.10.9

4. On other 2 nodes, run kubeadmin join to join the cluster.

kubectl get no -o wide

NAME      STATUS   ROLES                  AGE     VERSION    INTERNAL-IP    EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION     CONTAINER-RUNTIME
node02    Ready    <none>                 6m27s   v1.20.11   192.168.34.3   <none>        Ubuntu 20.04.3 LTS   5.4.0-88-generic   docker://20.10.9
node03    Ready    <none>                 5m6s    v1.20.11   192.168.34.4   <none>        Ubuntu 20.04.3 LTS   5.4.0-88-generic   docker://20.10.9
primary   Ready    control-plane,master   22m     v1.20.11   192.168.34.2   <none>        Ubuntu 20.04.3 LTS   5.4.0-88-generic   docker://20.10.9

Reference

1. https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/install-kubeadm/
2. https://www.packer.io/docs/builders/vagrant
3. https://www.vagrantup.com/docs/providers/virtualbox/networking
4. https://github.com/easzlab/kubeasz
5. https://www.tkng.io