Readiness Check

Problem

Your application process is running. But it's still loading data and cannot serve requests yet. Running does not necessarily mean ready for traffic.

Solution — Readiness Check

Infrastructure periodically sends an HTTP GET to /ready or /health, configured to check readiness. The app returns 200 when ready to serve requests or 503 while unavailable. After configured failure thresholds, Kubernetes stops routing Service traffic to the Pod until it becomes ready again. Process alive ✓ Ready ✗ 30 seconds later: Process alive ✓ Ready ✓

Core idea: readiness tells infrastructure whether a running workload should receive traffic.