https://drive.google.com/file/d/1ovRvqFddKp74eItuM_PUL61xBU4vIzsV/view?usp=sharing

🔍 YAML Breakdown

apiVersion: v1
kind: Pod
metadata:
  name: nnwebserver
spec:
  containers:
    - name: nnwebserver
      image: lovelearnlinux/webserver:v1
      livenessProbe:
        exec:
          command: ["cat", "/var/www/html/index.html"]  # ← Check if file exists
        initialDelaySeconds: 10   # ← Wait 10s before first check
        timeoutSeconds: 3         # ← Fail if check takes >3s
        periodSeconds: 20         # ← Repeat every 20s
        failureThreshold: 3       # ← Fail 3 times → restart container
      resources:
        requests:
          cpu: "500m"
          memory: "128Mi"
        limits:
          cpu: "1000m"
          memory: "256Mi"
      ports:
        - containerPort: 80

🔑 Key Insight:

This probe restarts the container if /var/www/html/index.html is missing or unreadable.

📌 How Liveness Probes Work

🎯 Failure Timeline:

• t=10s: First probe
• If fails → t=30s: Second probe
• If fails → t=50s: Third probe
• If fails → container restarted at t=50s

💡 Probe Types:

• exec: Run command (your example)
• httpGet: HTTP request (e.g., GET /health)
• tcpSocket: Open TCP connection (for non-HTTP apps)

🧪 k3s Lab: Break the App & Watch Auto-Restart

🔧 Step 1: Deploy the Pod

# Apply Pod
kubectl apply -f pod-with-health-check.yml

# Verify it's running
kubectl get pods nnwebserver

🔧 Step 2: Simulate App Failure

💡 Delete the critical file that the probe checks:

# Remove index.html (breaks the app)
kubectl exec nnwebserver -- rm /var/www/html/index.html

🔧 Step 3: Watch Kubernetes Heal Itself

# Watch Pod restarts
kubectl get pods nnwebserver -w

# ✅ Expected:
# NAME          READY   STATUS    RESTARTS   AGE
# nnwebserver   1/1     Running   0          10s
# nnwebserver   1/1     Running   1          45s  ← Restarted!

# Check logs to confirm
kubectl describe pod nnwebserver | grep -A 5 "Liveness"
# Events:
#   Warning  Unhealthy  50s  kubelet  Liveness probe failed: cat: /var/www/html/index.html: No such file or directory
#   Normal   Killing    50s  kubelet  Container nnwebserver failed liveness probe, will be restarted

🔍 Why did it restart?

• Probe failed 3 times (failureThreshold: 3)
• Kubernetes killed and restarted the container
• New container recreates /var/www/html/index.html (assuming image does this on startup)

🔧 Step 4: Clean Up