Deploy the Traefik charm

As part of a Kubernetes deployment, Traefik provides reverse proxy capabilities to ingress-requiring applications. In this tutorial, we’ll go step-by-step through the process of deploying and integrating the Traefik charm to set up a reverse proxy for a simple Flask application.

What you’ll do

  1. Deploy the Traefik charm

  2. Deploy a simple Flask application

  3. Integrate Traefik and the Flask application

  4. Verify the routing in the terminal and in a browser

What you’ll need

You will need a working station, e.g., a laptop, with AMD64 architecture. Your working station should have at least 4 CPU cores, 8 GB of RAM, and 30 GB of disk space.

Tip

You can use Multipass to create an isolated environment by running:

multipass launch 24.04 --name charm-tutorial-vm --cpus 4 --memory 8G --disk 30G

To be able to work inside the Multipass VM, log in with the following command:

multipass shell charm-tutorial-vm

This tutorial requires the following software to be installed on your working station (either locally or in the Multipass VM):

  • Juju 3.6+

  • Canonical Kubernetes 1.32+

Use Concierge to set up Juju and Canonical Kubernetes. Install Concierge with:

sudo snap install --classic concierge

For Traefik to receive an external IP address, the Kubernetes cluster must have a configured load balancer. Concierge should configure the load balancer for you, but we’ll use a custom configuration to set the IP of the load balancer to the IP of our working station or Multipass VM:

VM_IP=$(hostname -I | awk '{print $1}')
cat << EOF > concierge.yaml
providers:
k8s:
    enable: true
    bootstrap: true
    bootstrap-constraints:
    root-disk: "5G"
    features:
    load-balancer:
        l2-mode: "true"
        cidrs: "$VM_IP/28"
    local-storage: {}
    network: {}
    ingress:

host:
snaps:
    aws-cli:
EOF

sudo concierge prepare -c concierge.yaml

Note

If you’re using Multipass for this tutorial, this custom configuration means we can avoid setting up additional routes later in the tutorial when we visit the Flask application in a browser.

Verify that the load balancer is configured by running sudo k8s status and checking for load-balancer: enabled in the output.

For this tutorial, Juju must be bootstrapped to a Canonical Kubernetes controller. Concierge should complete this step for you, and you can verify by checking for msg="Bootstrapped Juju" provider=k8s in the terminal output and by running juju controllers.

If Concierge did not perform the bootstrap, run:

juju bootstrap k8s tutorial-controller

Set up the Juju model

To manage resources effectively and to separate this tutorial’s workload from your usual work, create a new model in the controller with:

juju add-model traefik-tutorial

Deploy the Traefik charm

Let’s begin by deploying the Traefik charm:

juju deploy traefik-k8s --trust

By default the latest stable release of the charm will be deployed. We must also use the --trust flag to provide Traefik with elevated permissions to interact with the Kubernetes environment.

The charm may need a couple of minutes to finish deploying. Monitor the status of the deployment with juju status. Once the deployment has finished, the output of juju status should look similar to:

ubuntu@charm-tutorial-vm:~$
juju status 
Model             Controller     Cloud/Region  Version  SLA          Timestamp
traefik-tutorial  concierge-k8s  k8s           3.6.25   unsupported  12:07:36Z

App          Version  Status  Scale  Charm        Channel        Rev  Address         Exposed  Message
traefik-k8s  2.11.49  active      1  traefik-k8s  latest/stable  377  10.152.183.205  no       Serving at http://10.114.45.129

Unit            Workload  Agent  Address     Ports  Message
traefik-k8s/0*  active    idle   10.1.0.75          Serving at http://10.114.45.129

Traefik is active, idle, and ready to route traffic. Now we need to provide Traefik with an application.

Deploy the Flask application

For this tutorial, we’ll use the Flask K8s charm as our application. Let’s deploy it now:

juju deploy flask-k8s --channel edge

Integrate Traefik and the Flask application

We’ll provide a communication pathway between the Flask application and Traefik by integrating them:

juju integrate traefik-k8s flask-k8s

Let’s check what’s going on with our deployment using juju status --relations:

ubuntu@charm-tutorial-vm:~$
juju status --relations 
Model             Controller     Cloud/Region  Version  SLA          Timestamp
traefik-tutorial  concierge-k8s  k8s           3.6.25   unsupported  12:09:52Z

App          Version  Status  Scale  Charm        Channel        Rev  Address         Exposed  Message
flask-k8s             active      1  flask-k8s    latest/edge     19  10.152.183.209  no
traefik-k8s  2.11.49  active      1  traefik-k8s  latest/stable  377  10.152.183.205  no       Serving at http://10.114.45.129

Unit            Workload  Agent  Address     Ports  Message
flask-k8s/0*    active    idle   10.1.0.229
traefik-k8s/0*  active    idle   10.1.0.75          Serving at http://10.114.45.129

Integration provider      Requirer                  Interface       Type     Message
flask-k8s:secret-storage  flask-k8s:secret-storage  secret-storage  peer
traefik-k8s:ingress       flask-k8s:ingress         ingress         regular
traefik-k8s:peers         traefik-k8s:peers         traefik_peers   peer

The key relation here is the one between Traefik and the Flask application:

Integration provider      Requirer                  Interface
traefik-k8s:ingress       flask-k8s:ingress         ingress

After we ran juju integrate traefik-k8s flask-k8s, Juju connected the two applications together through the ingress relation endpoint. The Flask application can now provide Traefik with its internal IP address and port (which is 8000 by default). In return, Traefik can now act as a reverse proxy for the application, automatically generating an external URL.

See also

ingress relation endpoint

We’ll now test whether the routing works.

Verify the routing

First, let’s verify that the Flask application serves traffic. We’ll need the IP address of the Flask unit listed in the output of juju status. In the example terminal output above, the IP address is 10.1.0.229. We can also grab this information generically using jq:

FLASK_IP=$(juju status --format json | jq -r '.applications."flask-k8s".units."flask-k8s/0".address')

Test the deployment using cURL:

curl $FLASK_IP:8000

If the deployment is successful, the output should show HTML containing <title>Welcome to flask-k8s Charm</title>.

Now we can check the external URL set up by Traefik to verify that traffic is routed through Traefik. To determine the external URL, let’s use the charm’s show-proxied-endpoints action:

juju run traefik-k8s/0 show-proxied-endpoints

The action lists all the endpoints proxied by Traefik. The command should output something similar to:

ubuntu@charm-tutorial-vm:~$
juju run traefik-k8s/0 show-proxied-endpoints 
Running operation 1 with 1 task
  - task 2 on unit-traefik-k8s-0

Waiting for task 2...
proxied-endpoints: '{"traefik-k8s": {"url": "http://10.114.45.129"}, "flask-k8s":
  {"url": "http://10.114.45.129/traefik-tutorial-flask-k8s"}}

Notice that Traefik has set up the URL http://10.114.45.129/traefik-tutorial-flask-k8s for our Flask application in this example output. Once again we’ll test with cURL:

curl http://10.114.45.129/traefik-tutorial-flask-k8s

The terminal should show the same HTML output as before, thus confirming that traffic is being routed through Traefik.

Visit in a browser

The HTML output from cURL can be difficult to read in a terminal. As a final step, let’s visit the Flask application in a browser.

Access the Flask application at http://10.114.45.129/traefik-tutorial-flask-k8s in your browser. You should see the message “Congratulations! You’ve successfully deployed the flask-k8s charm.”

Clean up the environment

Congratulations! You successfully deployed the Traefik charm, integrated it with a basic Flask application, and verified that the routing works by accessing the external URL.

You can clean up your environment by following this guide: Tear down your deployment

Next steps

You achieved a basic deployment of the Traefik charm. If you want to go farther in your deployment or learn more about the charm, check out these pages: