michalgutowski/cka-lab

Virtualbox lab environment for LFS258 Kubernetes Fundamentals training and practicing for CKA exams

The Vagrant file creates following Ubuntu (Xenial) VMs - master(s), worker and load-balancer.
Master nodes have been prepared for installing k8s cluster with kubeadm commands using kubernetes\kubeadm-config.yaml file Load-balancer node has been prepared with:

simple haproxy configuration to test external access for k8s services exposed with nodePort
simple nfs export /opt/sfw/ for testing persistent volumes

Simpe lab diagram

Cluster networking:

Internal IP addresses (you can copy the following to your /etc/hosts file):

192.168.56.100 ckalb  
192.168.56.101 ckamaster1 ckamaster  
192.168.56.102 ckamaster2  
192.168.56.103 ckamaster3  
192.168.56.104 ckaworker1

Pod CIDR:

10.244.0.0/16 (with Calico via enp0s8 interface)

Building clusters

For most labs one master + one worker + additional lb/nfs node is enough. To initially build the cluster execute:

# vagrant up ckamaster1 ckaworker1 ckalb

All VMs will be created under CKA group in your Virtualbox.
Once the cluster is up and running you can suspend it using:

# vagrant suspend ckamaster1 ckaworker1 ckalb

And later resume it whenever needed:

# vagrant resume ckamaster1 ckaworker1 ckalb

For the master HA labs you will need to build three masters + lb node

# vagrant up ckamaster1 ckamaster2 ckamaster3 ckalb

Connecting to cluster nodes

You can connect to all nodes using student user and submiting the private key e.g.:

# ssh student@192.168.56.101 -i id_rsa

Installing k8s master

In order to install first kubernetes master node connect to ckamaster1 and execute:

$ sudo -i
# apt install docker.io kubeadm=1.15.1-00 kubectl=1.15.1-00 kubelet=1.15.1-00
# kubeadm init --config=kubeadm-config.yaml --upload-certs| tee kubeadm-init.out

Now it's your turn to figure how to add additional worker or master nodes. Have fun!

Useful kubectl commands that helps you with creating objects

Creating a deployment
$ kubectl run

Creating a pod
$ kubectl run --restart=Never

Creating a job
$ kubectl run --restart=OnFailure

Creating a cronjob
$ kubectl run --restart=OnFailure --schedule=“* * * * *”

Notes

Passwordless ssh has been configured between nodes using root user
Password for user student: welcome1

Tested on MacOS 10.14.6 (Mojave), Vagrant 2.2.5 and VirtualBox 6.0.8, Ubuntu/Xenial vagrant box v. 20191114.0.0.
Min. 16GB of RAM recommended.

michalgutowski / cka-lab