Kubernetes: Learning the Hard Way

Welcome! This project is for learning Kubernetes (k8s) by doing everything manually. Below you'll find essential concepts and the most useful kubectl commands for hands-on learning.

Key Kubernetes Concepts

• Cluster: The whole Kubernetes system, made up of nodes.
• Node: A machine (virtual or physical) that runs your application workloads. A cluster has at least one node (often many).
• Pod: The smallest deployable unit in Kubernetes. A pod can hold one or more containers (usually one), and runs on a node.
• Deployment: Manages pods and ensures the right number are running at all times.
• Service: Exposes your pods to the network (internal or external), acting as a load balancer or gateway.

Step-by-Step: Recreate and Test Your Cluster

1. Create a new cluster:

   k3d cluster create cluster-1 -p "8081:80@loadbalancer"

   Uses your custom kind config to set up the cluster.

2. Install the metrics server (for HPA):

   kubectl apply -f metrics-server.yaml

   Enables resource metrics required for autoscaling.

3. Deploy your app and service:

   kubectl apply -f deployment.yaml

   Deploys the sample-app Deployment and Service.

4. Enable autoscaling with HPA:

   kubectl apply -f hpa.yaml

   Sets up the Horizontal Pod Autoscaler for your app.

5. (Optional) Deploy the load generator:

   kubectl apply -f load-generator.yaml

   Creates a pod that continuously sends requests to your app to trigger scaling.

6. (Optional) Deploy the Kubernetes Dashboard:

   kubectl apply -f kubernetes-dashboard.yaml

   Installs a web UI to view and manage your cluster.

   kubectl proxy

   kubectl port-forward -n kubernetes-dashboard service/kubernetes-dashboard 8001:443
   

   Starts a proxy server to access the dashboard.

   kubectl apply -f dashboard-admin.yaml

   Grants admin permissions to the dashboard.

---

7. (Optional) Deploy the Ingress:

   kubectl apply -f ingress.yaml

   Creates an Ingress resource to expose your app.

Useful kubectl Commands

• kubectl get nodes — List all nodes in your cluster
• kubectl get pods — List all pods in your cluster
• kubectl get deployments — List all deployments
• kubectl get services — List all services
• kubectl get hpa — List all Horizontal Pod Autoscalers
• kubectl describe pod <pod-name> — Detailed info about a pod
• kubectl logs <pod-name> — Show logs from a pod
• kubectl exec -it <pod-name> -- /bin/sh — Open a shell inside a running pod
• kubectl get all — List all resources (pods, services, deployments, etc.)
• kubectl delete -f <file.yaml> — Delete resources defined in a YAML file
• kubectl delete pod <pod-name> — Delete a specific pod

---

Load Generator

• kubectl apply -f load-generator.yaml — Apply the load generator pod
• kubectl exec -it load-generator -- /bin/sh — Open a shell inside the load generator pod
• kubectl delete -f load-generator.yaml — Delete the load generator pod

Clean Up

• kind delete cluster — Delete the entire cluster and all resources

---

Happy learning! Add more commands or notes here as you progress.

Tips

• Always check the status of your resources with kubectl get ....
• Use kubectl describe ... for detailed troubleshooting.
• Clean up with kubectl delete ... to avoid resource clutter.

Dashboard

• Access the dashboard at https://localhost:8001
• Use the token from kubectl -n kubernetes-dashboard create token admin-user