In our previous step we deployed a web app to Kubernetes on our Stack of NUCs.

Now we are going to install HAProxy.

Purpose:

• Demonstrate a using HAProxy in front of a web application on Kubernetes

References

• https://28gauravkhore.medium.com/how-to-configure-the-haproxy-using-the-ansible-and-also-how-to-configure-haproxy-dynamically-f18a55de3a66

From NUC 1, log in to the Ansible control node, NUC 2.

1. Create folder /home/ansible/my-project/haproxy and change directory to it
2. Copy the hosts file we created early in the lab
   • cp /home/ansible/my-project/hosts /home/ansible/my-project/haproxy/
3. Create the file /home/ansible/my-project/haproxy/ansible.cfg

   • ansible.cfg

     [defaults]
     inventory         = hosts

4. Create the file /home/ansible/my-project/haproxy/haproxy.cfg.j2

   • haproxy.cfg.j2

     # ------------------------
     # main frontend which proxys to the backends
     # ------------------------
     frontend main
         bind *:8080
         acl url_static  path_beg    -i /static /images /javascript /stylesheets
         acl url_static  path_end    -i .jpg .gif .png .css .js
      
         use_backend static  if url_static
         default_backend app
     # ------------------------
     # static backend for serving up images, stylesheets and such
     # ------------------------
     backend static
         balance roundrobin
         server  static 127.0.0.1:4331 check
     # ------------------------
     # round robin balancing between the various backends
     # ------------------------
     backend app
         balance roundrobin
     {% for ip in groups['nodes'] %}
         server app{{ loop.index}} {{ ip }}:80 check
     {% endfor %}

haproxy-install.yml

---
- hosts: nodes
  tasks:
    - name: Install haproxy
      package:
        name: haproxy
        state: present
    - name: Configure haproxy.cfg file
      template:
        src: "/home/ansible/my-project/haproxy/haproxy.cfg.j2"
        dest: "/etc/haproxy/haproxy.cfg"
    - name: "haproxy service start"
      service:
        name: haproxy
        state: restarted

ansible-playbook haproxy-install.yml

1. Point your browser to http://IPANYK8WORKER:8080/
2. Let's prove that any of the nodes will serve the app from any pod on any host
   • Change directory to /home/ansible/my-project/k8s
   • Edit speedtester-deployment.yml and change the number of replicas to 1
   • kubectl apply -f speedtester-deployment.yml
   • Use kubectl get pods to watch the pods terminated until 1 is left running
   • Point your browser to any of the k8s working IPs on port 8080 and note the app is still up
   • Set the replicas back to equal the number of worker nodes and apply the file again

You have completed this Lab! What did you think of it? What suggestions do you have for improvement? Let UncleNUC know!

A possible future lab could be written to install a ML model on a Kubernetes cluster so it can be applied to incoming data.

References:

• https://www.analyticsvidhya.com/blog/2022/01/deploying-ml-models-using-kubernetes/
• https://www.seldon.io/deploying-machine-learning-models-on-kubernetes