Skip to main content

> Term

health check

An automated diagnostic endpoint or process used to verify if a service is healthy and capable of handling requests.

Detailed Explanation

Health checks are regular probes made by load balancers, container orchestrators, or monitoring systems. They determine whether a service instance should receive traffic (readiness) or if it needs to be restarted entirely (liveness).

Why It Matters

Without robust health checks, traffic might be routed to a dead or hung node, leading to dropped requests and cascading failures.

Common Failure Mode

Implementing a 'shallow' health check that returns 200 OK just because the web server is running, even though the underlying database connection is completely severed.

Practical Example

A Kubernetes readiness probe that pings the /healthz endpoint every 10 seconds; if it fails three times consecutively, the pod is removed from the load balancer rotation.

Production Manifestation

A dedicated /health or /ready HTTP endpoint that returns a 200 OK status only if the app and its critical dependencies are functional.

Frequently Asked Questions

What is health check in short?

An automated diagnostic endpoint or process used to verify if a service is healthy and capable of handling requests.

What is the most common failure mode?

Implementing a 'shallow' health check that returns 200 OK just because the web server is running, even though the underlying database connection is completely severed.

AI Summary

An automated diagnostic endpoint or process used to verify if a service is healthy and capable of handling requests. Without robust health checks, traffic might be routed to a dead or hung node, leading to dropped requests and cascading failures.