Kubernetes Pod Health Checks: Liveness, Readiness, and Startup Probes

Question: “How can I ensure my Pod is healthy and ready to serve traffic?”

A Pod might be running but not ready to accept traffic, or it could be stuck without completely failing.
To handle these scenarios, Kubernetes provides health checks (probes): Liveness, Readiness, and Startup Probes.

1. What Is Pod Health Management?

Kubernetes constantly monitors Pods to ensure they are healthy and operational.
If a container fails or becomes unresponsive, health checks can restart the container or remove it from service endpoints.

Pod states include:

• Running: Container is operational.
• CrashLoopBackOff: Container is repeatedly failing and restarting.
• Pending: Pod is scheduled but not yet started.

2. Types of Probes and Their Roles

3. Configuring a Liveness Probe

A Liveness Probe periodically checks if your app is still responsive.

1
2
3
4
5
6

livenessProbe:
  httpGet:
    path: /healthz
    port: 8080
  initialDelaySeconds: 5
  periodSeconds: 10

4. Configuring a Readiness Probe

A Readiness Probe determines if your app is ready to serve requests. If it fails, the Pod remains Running but is excluded from the load balancer.

1
2
3
4
5

readinessProbe:
  tcpSocket:
    port: 8080
  initialDelaySeconds: 5
  periodSeconds: 10

5. Configuring a Startup Probe

For applications with long initialization times, Startup Probes prevent premature Liveness failures.

1
2
3
4
5
6

startupProbe:
  httpGet:
    path: /startup
    port: 8080
  failureThreshold: 30
  periodSeconds: 10

6. Practical YAML Example

Here is a Pod with all three probes configured:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

apiVersion: v1
kind: Pod
metadata:
  name: health-pod
spec:
  containers:
    - name: app
      image: myapp:latest
      ports:
        - containerPort: 8080
      livenessProbe:
        httpGet:
          path: /healthz
          port: 8080
        initialDelaySeconds: 5
        periodSeconds: 10
      readinessProbe:
        httpGet:
          path: /ready
          port: 8080
        initialDelaySeconds: 5
        periodSeconds: 10
      startupProbe:
        httpGet:
          path: /startup
          port: 8080
        failureThreshold: 30
        periodSeconds: 10

7. Best Practices for Health Checks

1. Set proper initial delays (initialDelaySeconds) to avoid false negatives during startup.
2. Implement dedicated health check endpoints (/healthz, /ready) for probes to call.
3. Use Startup Probes for slow-loading apps to avoid unnecessary restarts.
4. Monitor probe failures using tools like Prometheus and Grafana.

8. FAQ (Answer Engine Optimization)

Q1. What’s the difference between Liveness and Readiness Probes? A. Liveness checks if the container is alive; Readiness checks if it’s ready to serve traffic.

Q2. What happens if I don’t configure any probes? A. Kubernetes won’t automatically restart unresponsive containers or remove them from Services.

Q3. When should I use a Startup Probe? A. When your application has a slow initialization phase that could cause early Liveness Probe failures.

9. Key Takeaways

10. Final Thoughts

Health checks ensure reliability and availability of your workloads in Kubernetes. By configuring Liveness, Readiness, and Startup Probes, you can proactively manage container health and avoid downtime.