DEPLOY A KUBERNETES CLUSTER USING ANSIBLE

To deploy a Kubernetes cluster on Linux using Ansible Playbooks we have found Ansible to be a fantastic tool for getting a Kubernetes cluster up and running quickly in my development environment, and now use the Ansible playbooks when we need to stand up a Kubernetes cluster quickly and easily.

We will use Ansible to deploy a small Kubernetes cluster – with one master node, used to manage the cluster, and two worker nodes, which will be used to run our container applications. To achieve this, we will use four Ansible playbooks. These will do the following:

  1. Create a new User Account for use with Kubernetes on each node

  2. Install Kubernetes and containerd on each node

3. Configure the Master node 3. Join the Worker nodes to the new cluster

If you are considering using Ansible to deploy Kubernetes already, we will assume you’re already somewhat familiar with both technologies. So, with that said, let’s get straight into the detail.

PREREQUISITES

Before we can get started, we need a few prerequisites to be in place. This is what we are going to need:

  1. A host with Ansible installed. I’ve written previously about how to install Ansible – also, check out the online documentation! You should also set up an SSH key pair, which will be used to authenticate to the Kubernetes nodes without using a password, allowing Ansible to do it’s thing.

  2. Three servers/hosts to which we will use as our targets to deploy Kubernetes. Our servers each have 12GB ram and 4vCPUs. This is fine for my lab purposes, which we use to try out new things using Kubernetes. You need to be able to SSH into each of these nodes as root using the SSH key pair we mentioned above.

With that lot all in place we should be ready to go!

Setting up Ansible to Deploy Kubernetes

Before we start to look at the Ansible Playbooks, we need to set up Ansible to communicate with the Kubernetes nodes. First of all, on our Ansible host, lets set up a new directory from which we we run our playbooks.

mkdir kubernetes
cd kubernetes

With that done, we now need to create a hosts file, to tell Ansible how to communicate with the Kubernetes master and worker nodes.

vi hosts

The content of the hosts file should look something like the following:

[masters]
master ansible_host=xxx.xxx.xxx.xxx ansible_user=root

[workers]
worker1 ansible_host=yyy.yyy.yyy.yyy ansible_user=root
worker2 ansible_host=zzz.zzz.zzz.zzz ansible_user=root

Note

  • x - Master IP Address

  • y - Worker1 IP Address

  • z - Worker2 IP Address

Listing the master node and the worker nodes in different sections in the hosts file will allow us to target the playbooks at the specfic node type later on.

Finally, with that done, we can test it’s working by doing a Ansible ping:

ansible -i hosts all -m ping
Output

master | SUCCESS => {
    "changed": false,
    "ping": "pong"
}
worker2 | SUCCESS => {
    "changed": false,
    "ping": "pong"
}
worker1 | SUCCESS => {
    "changed": false,
    "ping": "pong"
}

All good! Lets move onto the first playbook.

Creating a Kubernetes user with Ansible Playbook

Our first task in setting up the Kubernetes cluster is to create a new user on each node. This will be a non-root user, that has sudo privileges. It’s a good idea not to use the root account for day to day operations, of course. We can use Ansible to set the account up on all three nodes, quickly and easily. First, create a file in the working directory:

vim users.yml

Then add the following to the playbook:

- hosts: 'workers, masters'
  become: yes

  tasks:
    - name: create the kube user account
      user: name=kube append=yes state=present createhome=yes shell=/bin/bash

     - name: allow 'kube' to use sudo without needing a password
       lineinfile:
            dest: /etc/sudoers
            line: 'kube ALL=(ALL) NOPASSWD: ALL'
            validate: 'visudo -cf %s'

     - name: set up authorized keys for the kube user
       authorized_key: user=kube key="{{item}}"
       with_file:
          - ~/.ssh/id_rsa.pub

We’re now ready to run our first playbook. To do so:

ansible-playbook -i hosts users.yml

Once done you should see:

../_images/A9.jpg

Install Kubernetes with Ansible Playbook

Now we’re getting to the fun part! With our user now created, we can move on to installing Kubernetes. Lets dive straight in and have a look at the playbook, which we have named install-k8s.yml:

Sample Playbook

- hosts: "masters, workers"
remote_user: ubuntu
become: yes
become_method: sudo
become_user: root
gather_facts: yes
connection: ssh

tasks:
    - name: Create containerd config file
    file:
        path: "/etc/modules-load.d/containerd.conf"
        state: "touch"

    - name: Add conf for containerd
    blockinfile:
        path: "/etc/modules-load.d/containerd.conf"
        block: |
            overlay
            br_netfilter

    - name: modprobe
    shell: |
            sudo modprobe overlay
            sudo modprobe br_netfilter


    - name: Set system configurations for Kubernetes networking
    file:
        path: "/etc/sysctl.d/99-kubernetes-cri.conf"
        state: "touch"

    - name: Add conf for containerd
    blockinfile:
        path: "/etc/sysctl.d/99-kubernetes-cri.conf"
        block: |
                net.bridge.bridge-nf-call-iptables = 1
                net.ipv4.ip_forward = 1
                net.bridge.bridge-nf-call-ip6tables = 1

    - name: Apply new settings
    command: sudo sysctl --system

    - name: install containerd
    shell: |
            sudo apt-get update && sudo apt-get install -y containerd
            sudo mkdir -p /etc/containerd
            sudo containerd config default | sudo tee /etc/containerd/config.toml
            sudo systemctl restart containerd

    - name: disable swap
    shell: |
            sudo swapoff -a
            sudo sed -i '/ swap / s/^\(.*\)$/#\1/g' /etc/fstab

    - name: install and configure dependencies
    shell: |
            sudo apt-get update && sudo apt-get install -y apt-transport-https curl
            curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -

    - name: Create kubernetes repo file
    file:
        path: "/etc/apt/sources.list.d/kubernetes.list"
        state: "touch"

    - name: Add K8s Source
    blockinfile:
        path: "/etc/apt/sources.list.d/kubernetes.list"
        block: |
            deb https://apt.kubernetes.io/ kubernetes-xenial main

    - name: install kubernetes
    shell: |
            sudo apt-get update
            sudo apt-get install -y kubelet=1.20.1-00 kubeadm=1.20.1-00 kubectl=1.20.1-00
            sudo apt-mark hold kubelet kubeadm kubectl

This playbook will run against all three nodes, and will install the containerd runtime (including some pre-requisite configuration, then go onto install Kubernetes, which includes kubelet, kubeadm and kubectl. Run the playbook using the following syntax:

ansible-playbook -i hosts install-k8s.yml

There’s quite a lot going on here, so this one will take a little while to run whist the necessary packages are installed on each node. Once done you should see:

PLAY RECAP ****************************************************************************
master                     : ok=13   changed=12   unreachable=0    failed=0
worker1                    : ok=13   changed=12   unreachable=0    failed=0
worker2                    : ok=13   changed=12   unreachable=0    failed=0

Now we’re halfway there!

Creating a Kubernetes Cluster Master Node using Ansible Playbook

Now we should have containerd and Kubernetes installed on all our nodes. The next step is to create the cluster on the master node. This is the master.yml file, which will initialise the Kubernetes cluster on my master node and set up the pod network, using Calico:

- hosts: masters
become: yes
tasks:
    - name: initialize the cluster
    shell: kubeadm init --pod-network-cidr=10.244.0.0/16
    args:
        chdir: $HOME
        creates: cluster_initialized.txt

    - name: create .kube directory
    become: yes
    become_user: kube
    file:
        path: $HOME/.kube
        state: directory
        mode: 0755

    - name: copies admin.conf to user's kube config
    copy:
        src: /etc/kubernetes/admin.conf
        dest: /home/kube/.kube/config
        remote_src: yes
        owner: kube

    - name: install Pod network
    become: yes
    become_user: kube
    shell: kubectl apply -f https://docs.projectcalico.org/manifests/calico.yaml
    args:
        chdir: $HOME

    - name: Get the token for joining the worker nodes
    become: yes
    become_user: kube
    shell: kubeadm token create  --print-join-command
    register: kubernetes_join_command

    - debug:
    msg: "{{ kubernetes_join_command.stdout }}"

    - name: Copy join command to local file.
    become: yes
    local_action: copy content="{{ kubernetes_join_command.stdout_lines[0] }}" dest="/tmp/kubernetes_join_command" mode=0777

Note, towards the end of the playbook we generate the worker join command and save it to a local file on the Ansible host. We will use this file later to join the worker nodes to the cluster. Before then, execute the masters.yml playbook:

ansible-playbook -i hosts master.yml

Once the playbook has finished, we can check the outcome by connecting to the cluster master node using SSH to check the status of the master node:

kubectl get nodes
Output

NAME                           STATUS   ROLES                  AGE   VERSION
kube01.test.local              Ready    control-plane,master   13m   v1.20.1

From the output, we can see that the master node’s status is Ready, showing that the cluster has been initialised successfully.

Join Worker Nodes to Kubernetes Cluster using Ansible Playbook

Now we have a Kubernetes cluster initialised, the final step is to join our worker nodes to the cluster. To do so, the final playbook – join-workers.yml – contains the following:

- hosts: workers
become: yes
gather_facts: yes

tasks:
- name: Copy join command from Ansiblehost to the worker nodes.
    become: yes
    copy:
    src: /tmp/kubernetes_join_command
    dest: /tmp/kubernetes_join_command
    mode: 0777

- name: Join the Worker nodes to the cluster.
    become: yes
    command: sh /tmp/kubernetes_join_command
    register: joined_or_not

This works by copying the file containing the worker join command saved locally earlier to the worker nodes, then it runs the command. Run the playbook with:

ansible-playbook -i hosts join-workers.yml

Once the playbook has complete we can check the status of the cluster nodes by again running the following on the cluster master node:

kubectl get nodes
Output

NAME                           STATUS   ROLES                  AGE   VERSION
kube02.test.local              Ready    <none>                 80s   v1.20.1
kube01.test.local              Ready    control-plane,master   22m   v1.20.1
kube03.test.local              Ready    <none>                 69s   v1.20.1