Demo Working with Kubernetes

In this lesson, you'll learn how to interact with Kubernetes clusters on AWS EKS using the kubectl CLI. We have two clusters set up—one for production and another for staging. Follow along as we inspect the kubeconfig file, switch between clusters, deploy a Flask application, perform load testing with k6, and update the application image.

Accessing the EKS Clusters

Open the AWS EKS console to view your clusters. In our setup, the console shows two active clusters named "prod" and "staging," both running Kubernetes version 1.27 with extended support until July 24, 2025.

To interact with these clusters, you need the kubectl CLI utility. Follow the installation instructions provided in the Kubernetes documentation tailored for your operating system.

Inspecting the kubeconfig File

Your kubeconfig file, usually located at ~/.kube/config, contains configurations for your clusters, users, and contexts. To inspect the file, run:

cat ~/.kube/config > config.yaml

Open config.yaml in your preferred IDE.

A section from the file for the staging cluster might look like this:

apiVersion: v1
clusters:
  cluster:
    certificate-authority: C:\Users\sanje\.minikube\ca.crt
    extensions:
      - extension:
          last-update: Sun, 04 Aug 2024 23:53:45 MDT
user:
  exec:
    apiVersion: client.authentication.k8s.io/v1beta1
    args:
      - eks
      - get-token
      - --output
      - json
      - --cluster-name
      - staging
      - --region
      - us-east-1
    command: aws
    env:
      - name: AWS_STS_REGIONAL_ENDPOINTS
        value: regional
    interactiveMode: IfAvailable
    provideClusterInfo: false

Similarly, the file includes configurations for the production cluster. Notice that the current-context is set to the production cluster by default:

contexts:
  - context:
      cluster: prod.us-east-1.eksctl.io
      user: [email protected]
      name: [email protected]
  - context:
      cluster: staging.us-east-1.eksctl.io
      user: [email protected]
      name: [email protected]
current-context: [email protected]
kind: Config
preferences: {}

Switching Between Clusters

To operate on different clusters, switch contexts. First, check the nodes on your current production cluster:

kubectl get node

The output might resemble:

NAME                           STATUS   ROLES    AGE   VERSION
ip-192-168-60-39.ec2.internal  Ready    <none>   5h33m v1.27.15-eks-1552ad0
ip-192-168-8-155.ec2.internal   Ready    <none>   5h33m v1.27.15-eks-1552ad0

To switch to the staging cluster, execute:

kubectl config use-context [email protected]

After switching, verify the change:

kubectl get node

The displayed node names will differ, confirming that you are now connected to the staging cluster.

Deploying the Flask Application

Our Kubernetes configurations are stored in the k8s/ directory, which includes the deployment.yaml and service.yaml files.

Deployment Configuration

The deployment.yaml file sets up a deployment for a Flask application with three replicas:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: flask-app
spec:
  replicas: 3
  selector:
    matchLabels:
      app: flask-app
  template:
    metadata:
      labels:
        app: flask-app
    spec:
      containers:
        - name: flask-app
          image: sanjeevkt720/jenkins-flask-app:v3
          resources:
            limits:
              memory: "128Mi"
              cpu: "500m"
          ports:
            - containerPort: 5000

Service Configuration

The service.yaml file defines a LoadBalancer service to expose the Flask application externally:

apiVersion: v1
kind: Service
metadata:
  name: flask-app-service
spec:
  type: LoadBalancer
  selector:
    app: flask-app
  ports:
    - port: 5000
      targetPort: 5000

This configuration assigns an external IP and possibly a DNS name that routes incoming traffic to the correct pods.

Deploying Resources to the Cluster

Before deploying, confirm that your current context is correct:

kubectl config current-context

Deploy all resources (both deployment and service) with:

kubectl apply -f k8s/

You should see output similar to:

deployment.apps/flask-app created
service/flask-app-service created

Next, verify that the deployment and pods are running:

kubectl get deployment

Example output:

NAME        READY   UP-TO-DATE   AVAILABLE   AGE
flask-app   3/3     3            3           177m

And check the pods:

kubectl get pod

Example output:

NAME                               READY   STATUS    RESTARTS   AGE
flask-app-795b844d7f-24pfw          1/1     Running   0          177m
flask-app-795b844d7f-7cdp           1/1     Running   0          177m
flask-app-795b844d7f-lmclx          1/1     Running   0          177m

Finally, retrieve the service details to access your application:

kubectl get svc

Example output:

NAME                  TYPE           CLUSTER-IP     EXTERNAL-IP                                                             PORT(S)
flask-app-service     LoadBalancer   10.100.67.6    a2718e7dae7964585baeb017221aeb6d-1943673624.us-east-1.elb.amazonaws.com   5000/TCP
kubernetes            ClusterIP      10.100.0.1     <none>                                                                  443/TCP

Copy the EXTERNAL-IP or hostname and access your application on port 5000.

Load Testing with k6

Use k6 to simulate user traffic and measure your application's performance. k6 lets you define virtual users (VUs), duration, and thresholds for request response times.

Create a file named acceptance-test.js with the following content:

import http from "k6/http";
import { sleep } from "k6";

export const options = {
  vus: 10,
  duration: "10s",
  thresholds: {
    http_req_duration: ["p(90)<600"], // 90% of requests must complete in under 600ms
  },
};

export default function () {
  http.get(`http://${__ENV.SERVICE}`);
  sleep(1);
}

The target URL is provided via an environment variable. Retrieve the external hostname and port of your service with a command similar to:

kubectl get svc flask-app-service -o jsonpath='{.status.loadBalancer.ingress[0].hostname}:{.spec.ports[0].port}'

Then run the k6 test by setting the SERVICE environment variable:

k6 run -e SERVICE=<external-hostname>:5000 acceptance-test.js

During the test, k6 will simulate 10 virtual users over 10 seconds, ensuring that 90% of the requests complete in under 600 milliseconds. After the test, you'll see statistical data including average, median, and maximum response times.

Updating the Application Version

Our current deployment uses version 3 of the image:

spec:
  template:
    metadata:
      labels:
        app: flask-app
    spec:
      containers:
        - name: flask-app
          image: sanjeevkt720/jenkins-flask-app:v3
          resources:
            limits:
              memory: "128Mi"
              cpu: "500m"
          ports:
            - containerPort: 5000

To update the application to a new version (for example, v5), follow these steps:

1. Build the New Image

   docker build -t sanjeevkt720/jenkins-flask-app:v5 .
   

2. Push the Image to Docker Hub

   docker push sanjeevkt720/jenkins-flask-app:v5
   

3. Update the Deployment in Kubernetes

   Use the kubectl set image command to update the image version:

   kubectl set image deployment/flask-app flask-app=sanjeevkt720/jenkins-flask-app:v5
   

   You should see a confirmation:

   deployment.apps/flask-app image updated
   

Finally, verify the rollout status of your deployment and refresh your application to ensure it is now using version 5.