Kubernetes Health Probes - Liveness vs Readiness¶

category: Kubernetes tags: health-checks, liveness, readiness, probes, k8s, monitoring

Main Topic 1¶

Kubernetes uses health probes to automatically detect when containers are alive and ready to serve traffic. This distinction is critical for high-availability and fault-tolerant systems.

Subtopic A: Types of Probes¶

Liveness Probe – Determines if a container is still running. If it fails, Kubernetes will restart the container.
Readiness Probe – Determines if a container is ready to accept traffic. If it fails, Kubernetes will remove the pod from the Service endpoints.

Subtopic B: Why the Difference Matters¶

Liveness keeps the container from running in a "zombie" state (e.g. stuck or deadlocked).
Readiness protects users from hitting an app that is still initializing or temporarily unavailable due to backend issues.

Main Topic 2¶

Example: Liveness Check (Lightweight)¶

// Lightweight liveness check - ensures app is running
app.get('/health/live', (req, res) => {
  res.status(200).json({ status: 'alive', timestamp: Date.now() });
});

This will return 200 as long as the server is up and event loop is alive.
It does NOT check whether the app is doing real work or connected to dependencies.

Example: Readiness Check (Deeper)¶

// Readiness check - ensure dependencies are available
app.get('/health/ready', async (req, res) => {
  try {
    await checkDatabaseConnection();
    await checkS3Availability();
    res.status(200).json({ status: 'ready' });
  } catch (err) {
    res.status(503).json({ status: 'unavailable', error: err.message });
  }
});

This returns 200 only when dependencies (DB, S3, etc.) are reachable.
Useful to gate real traffic until the app is fully operational.

Key Concepts Summary¶

Liveness Probe – Keeps the app from getting stuck or dead.
Readiness Probe – Prevents traffic from being routed to an unready pod.
Status 200 ≠ healthy unless it reflects all required services.
Keep liveness light; readiness can be deep.

Probe Comparison Summary¶

Probe Type	What It Checks	Result if Failing	Example Use
Liveness	Is the app process running?	Container is restarted	Memory leak, deadlock
Readiness	Is the app ready to serve traffic?	Pod removed from endpoints	Waiting for DB, warmup

Best Practices / Tips¶

Never include external checks in liveness probes.
Use readiness probes to delay traffic until app is ready.
Separate logic clearly — don’t reuse /health for both purposes.
Use different paths and functions for /health/live vs /health/ready.
Set meaningful initial delays and timeouts based on startup requirements.