Kubernetes Health Probes: How Liveness, Readiness, and Startup Probes Work

Kubernetes offers three types of health probes — liveness, readiness, and startup — each serving a distinct purpose in managing container lifecycle. The liveness probe monitors whether a container is functioning correctly and triggers a restart if it fails, but should only check internal process health, not external dependencies like databases. The readiness probe determines whether a pod should receive traffic and is the appropriate place to check external dependencies, also playing a key role in ensuring safe rolling updates. Introduced in Kubernetes 1.18, the startup probe handles slow-initializing applications by disabling liveness and readiness checks until the app has fully started. Misconfiguring or omitting these probes can lead to crash loops, dropped traffic, or failed deployments, while correct configuration enables self-healing and zero-downtime rollouts.