Build a Kubernetes Cluster

Each of the following commands need to be run on all three servers unless otherwise specified.

1. From the jumphost using putty open a new SSH session to each of the following servers. Putty sessions are pre-configured to connect with the default user "ubuntu" and cert.

   • kube-master
   • kube-node1
   • kube-node2

2. Connect as root

   $ su -
   $ passwd = default
   

3. Edit /etc/hosts and add the following static host entries

   10.1.20.21    kube-master
   10.1.20.22    kube-node1
   10.1.20.23    kube-node2
   

4. Upgrade Ubuntu to ensure an up-to-date OS

   $ apt update && apt upgrade -y
   

5. Add the docker repo

   $ curl \-fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add \-
   $ add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"
   

6. Add the kubernetes repo

   $ curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add -
   $ cat <<EOF > /etc/apt/sources.list.d/kubernetes.list
   deb http://apt.kubernetes.io/ kubernetes-xenial main
   EOF
   

7. Install docker

   $ apt update && apt install -y docker-ce
   

8. Configure docker to use the correct cgroupdriver

   $ cat << EOF > /etc/docker/daemon.json
   {
   "exec-opts": ["native.cgroupdriver=cgroupfs"]
   }
   EOF
   

9. Verify docker is up and running? (should see the hello-world container pulled and ran with a "hello world" message.)

   $ docker run hello-world
   

10. Install kubernetes

    $ apt install -y kubelet kubeadm kubectl
    

11. Initialize kubernetes with default network. (default flannel network 10.244.0.0/16)

    Note

    Master ONLY!

    $ kubeadm init --pod-network-cidr=10.244.0.0/16
    

    Note

    ...the output. It will be needed to join the nodes to the master in a later step.

12. Configure kubernetes management. At this point you should be logged in as root. The following will update both root and ubuntu user accounts.

    Note

    Master ONLY!

    $ mkdir -p $HOME/.kube
    $ sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
    $ sudo chown $(id -u):$(id -g) $HOME/.kube/config
    $ logout
    $ mkdir -p $HOME/.kube
    $ sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
    $ sudo chown $(id -u):$(id -g) $HOME/.kube/config
    

13. Install flannel on the master. (default flannel network 10.244.0.0/16)

    Note

    Master ONLY!

    $ kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
    

14. Verify Kubernetes is up and running? (should see several kubernetes pods up and running.)

    $ kubectl get pods --all-namespaces
    

    Note

    Before running next step wait for all system pods to show status "Running"

15. Add the kubernetes nodes to the cluster. (cut and past the command from the previous "kubeadm init" output. It will look something like this...

    Note

    Node 1 & 2 ONLY!

    $ kubeadm join --token 7f92b3... 10.1.20.21:6443 --discovery-token-ca-cert-hash sha256:9c4...
    

16. Verify kube-node 1 & 2 are up and running

    $ kubectl get nodes