What You’ll Learn

• Introduction to Terraform
• Configure Terraform to use with Kubernetes
• Deploy resources to the Kubernetes cluster

What You’ll Need

• Deploy a Kubernetes cluster
• Install Terraform
• Grab VS Code or an equivalent editor
• Clone this GitHub repository
• Have fun!

• Terraform is a popular open-source infrastructure-as-code tool:
  • Primarily maintained by HashiCorp
  • Debuted in 2014
  • Uses a custom language called HashiCorp Configuration Language (HCL)
• Terraform manages resources using a pluggable component called a provider:
  • Providers are available for a wide variety of clouds and infrastructure vendors.
  • HashiCorp maintains a public provider registry for easy discovery and reuse of providers.

Core Terraform Commands and Workflow

• By convention, a Terraform configuration comprises three files:
  • variables.tf: Declares variables used to parameterize the configuration
  • main.tf: Describes the infrastructure to be provisioned declaratively, making frequent references to a variety of providers
  • output.tf: Declares output values that are displayed upon instantiation of the declared infrastructure and can then be referenced by other tooling

• Here’s a typical command flow using the Terraform CLI:
  • terraform init: Initializes the working directory containing the *.tf files, including the creation of a state file, and downloads off any referenced providers
  • terraform plan: Creates and displays an execution plan based on the infrastructure declared in the *.tf files
  • terraform apply: Executes the plan based on the infrastructure declared in the *.tf files and updates the state as pieces of the declaration are instantiated
  • terraform destroy: Deletes all the infrastructure declared in the *.tf files and updates the state accordingly

Now that we’ve covered the fundamentals of how we get started with Terraform, let’s leverage it to:

1. Deploy some NGINX nodes on top of our Kubernetes cluster.
2. Add a NodePort service and scale our deployment of NGINX.
3. Destroy all resources.

Simple enough? Let’s get rollin'!

Before we start, let’s make sure we have cloned our project 👉🏽 This GitHub repository

git clone https://github.com/CiscoLearning/deploy-terraform-kubernetes.git

1. Navigate to the directory deploy-terraform-kubernetes.

In this working directory, there are three relevant files (for now):

• kubernetes_resources.tf: This file contains the declared resources, two NGINX pods.
• kubernetes.tf: This file contains our main Terraform configuration file.
• terraform.tfvars: This file contains all our variables that we need to define for Terraform to successfully connect to our Kubernetes instance.

2. Head over into terraform.tfvars. Here’s where we need to update our variables with your own Kubernetes cluster information:

host                   = "REPLACE THIS"
client_certificate     = "REPLACE THIS"
client_key             = "REPLACE THIS"
cluster_ca_certificate = "REPLACE THIS"

Gather the variable information from our Kubernetes clusters, open a Terminal, and execute:

kubectl config view --minify --flatten --context=kind-my-kubernetes-cluster

This will return the information we need to update our Terraform var file. The output should look something like this:

3. Let’s update our terraform.tfvars file with the information gathered above. Copy the information from your Terminal output and paste in place of “REPLACE THIS”. Here’s how it maps:

• host :: server
• client_certificate :: client-certificate-data
• client_key :: client-key-data
• cluster_ca_certificate :: certificate-authority-data

Your terraform.tfvars file should now look similar to this:

At this point, we are ready to deploy our nodes using Terraform! Do you remember the Terraform flow?

In your Terminal, execute the following in order:

1. Initialize your working directory:

terraform init

2. Plan your resources deployment to your Kubernetes cluster:

terraform plan

3. Deploy your resources to your Kubernetes cluster; confirm the changes by typing “yes” and Enter:

terraform apply

🎉 We just deployed our first node service on top of our cluster! 🎉

Terraform is stateful, which means it knows exactly what is happening with our infrastructure at a given point—which also means we can update our infrastructure as needed. Pretty cool, right?

Let’s make our NGINX deployment available to the outside world by deploying a NodePort service and updating our configuration.

• In your working directory, check out kubernetes_service.tf.hold.

⚠️ Before you proceed any further, rename the file to kubernetes_service.tf by removing the trailing .hold from the filename.

• kubernetes_service.tf: This file contains the declaration of our NodePort service.

Add the service!

1. Check out what the new plan looks like before deployment:

terraform plan

2. Deploy your resources to your Kubernetes cluster; confirm the changes by typing “yes” and Enter:

terraform apply

What if our NGINX deployment is seeing high utilization that’s degrading the UX of our application, and we need to scale up the node? It’s not an issue; Terraform has you covered.

Let’s Scale Up the Node

1. Check out and edit the kubernetes_resources.tf file.
2. Modify the replicas from 2 to 4:

spec {
    replicas = 4			 # Change this value from 2 to 4
    selector {
      match_labels = {
        App = "longlivethebat"
      }
    }

3. Apply your changes and confirm that the replicas were changed from 2 to 4, then type “yes.” Enter a value line to confirm the apply:

terraform apply

4. Verify your deployment is now using 4 pods:

kubectl get deployments

As Andrea Bocelli might say, it’s “Time to Say Goodbye” to your cluster!

⚠️ If you’d like to keep your deployment running for further learning, skip this step.

1. Destroy your resources, including your NGINX deployment and NodePort service:

terraform destroy

2. On the Enter a value line, type “yes” to confirm the destroy. The destroy takes a moment to complete.
3. Delete your cluster:

kind delete cluster --name my-kubernetes-cluster

4. Verify the cluster was deleted:

kind get clusters

Congratulations! You’ve successfully completed this lab!

Learn More

• Visit Cisco U. Today
• Cisco Learning Network DevNet Certifications Community
• DevNet Certification Training Videos
• Train Your Team with Automation Bootcamps