Liveness probe concept in Kubernetes - Commands & Configuration

Introduction

Sometimes, applications inside containers can stop working but still appear running. A liveness probe helps Kubernetes check if an app is alive and restart it if it is stuck or crashed.

When your app might freeze or get stuck and needs a restart to recover.

When you want Kubernetes to automatically fix apps that stop responding.

When running long-lived services that must stay healthy without manual checks.

When you want to avoid manual intervention for crashed or deadlocked containers.

When you want to improve app reliability by automatic self-healing.

Config File - pod.yaml

pod.yaml

apiVersion: v1
kind: Pod
metadata:
  name: liveness-demo
spec:
  containers:
  - name: my-app
    image: nginx:1.23
    livenessProbe:
      httpGet:
        path: /healthz
        port: 80
      initialDelaySeconds: 5
      periodSeconds: 10
      timeoutSeconds: 1
      failureThreshold: 3

This file creates a pod named liveness-demo running an Nginx container.

The livenessProbe section tells Kubernetes to check the /healthz URL on port 80 every 10 seconds, starting 5 seconds after the container starts.

If the check fails, Kubernetes restarts the container automatically.

Commands

This command creates the pod with the liveness probe defined. Kubernetes will start the container and begin health checks.

Terminal

kubectl apply -f pod.yaml

Expected Output

pod/liveness-demo created

This command shows the status of all pods, including our new pod, to verify it is running.

Terminal

kubectl get pods

Expected Output

NAME           READY   STATUS    RESTARTS   AGE

liveness-demo  1/1     Running   0          10s

This command shows detailed information about the pod, including events related to the liveness probe and any restarts.

Terminal

kubectl describe pod liveness-demo

Expected Output

Name:         liveness-demo

Namespace:    default

Containers:

  my-app:

    Container ID:   docker://abcdef123456

    Image:          nginx:1.23

    State:          Running

    Ready:          True

    Restart Count:  0

    Liveness:       http-get http://:80/healthz delay=5s timeout=1s period=10s #success=1 #failure=3

Events:

  Type    Reason     Age   From               Message

  ----    ------     ----  ----               -------

  Normal  Scheduled  1m    default-scheduler  Successfully assigned default/liveness-demo to node-1

Key Concept

If you remember nothing else from this pattern, remember: a liveness probe lets Kubernetes detect and fix stuck or crashed containers automatically by restarting them.

Common Mistakes

Not setting initialDelaySeconds causing the probe to run before the app is ready

The probe fails immediately and Kubernetes restarts the container repeatedly, causing a crash loop.

Set initialDelaySeconds to a value that gives your app time to start before health checks begin.

Using a wrong path or port in the httpGet probe

The probe always fails because it cannot reach the correct endpoint, causing unnecessary restarts.

Verify the path and port match your app's health endpoint exactly.

Confusing livenessProbe with readinessProbe

Liveness restarts the container; readiness controls traffic routing. Using the wrong probe can cause unwanted restarts or traffic issues.

Use livenessProbe to check if the app is alive and restart if needed; use readinessProbe to control if the pod receives traffic.

Summary

Define a liveness probe in your pod spec to let Kubernetes check if your app is alive.

Use kubectl apply to create the pod and kubectl get pods to verify it is running.

Use kubectl describe pod to see detailed probe status and container restarts.