Kubernetes health checks: liveness vs readiness vs startup

Need to configure proper health checks so Kubernetes can detect and handle unhealthy pods correctly.

Three types of probes, each with a different purpose:

apiVersion: v1
kind: Pod
spec:
  containers:
  - name: app
    image: myapp:1.0
    ports:
    - containerPort: 8080
    
    # Startup probe: Is the app ready to start serving?
    # Disables liveness/readiness until it succeeds
    startupProbe:
      httpGet:
        path: /health
        port: 8080
      failureThreshold: 30    # 30 * 10s = 5 min max startup
      periodSeconds: 10
    
    # Liveness probe: Is the app still alive?
    # Failure = restart the container
    livenessProbe:
      httpGet:
        path: /health
        port: 8080
      initialDelaySeconds: 0   # startupProbe handles delay
      periodSeconds: 10
      failureThreshold: 3     # 3 failures = restart
    
    # Readiness probe: Can the app handle traffic?
    # Failure = remove from Service endpoints (no traffic)
    readinessProbe:
      httpGet:
        path: /ready
        port: 8080
      periodSeconds: 5
      failureThreshold: 2

When each fires:

• Startup: Once at boot. Blocks other probes until success
• Liveness: Continuously. Detects deadlocks/hangs -> restart
• Readiness: Continuously. Detects temporary inability -> stop traffic

Common health endpoint:

@app.get('/health')
def health():
    return {'status': 'ok'}  # App is alive

@app.get('/ready')
def ready():
    if not db.is_connected():
        return Response(status_code=503)
    return {'status': 'ready'}  # App can handle requests

Without proper probes, K8s can't distinguish between a slow startup, a temporary issue, and a dead process. Wrong probe type = wrong recovery action.

Kubernetes deployments needing proper health checking