Status API Readiness Lag
warningKong's readiness endpoint doesn't immediately respond 200 OK after startup because it takes time to load configuration and build data structures. Load balancers using this probe may prematurely route traffic to unready nodes.
Monitor time between Kong process start and first successful readiness probe response. Track 503 errors immediately after pod startup or rolling updates. If load balancers forward traffic before Kong responds 200 to /status readiness endpoint, traffic will fail.
Configure load balancers and Kubernetes probes with sufficient initialDelaySeconds and periodSeconds to account for Kong startup time. Don't expect immediate 200 OK—allow 10-30 seconds for configuration loading. Use liveness probe separately from readiness to distinguish node health from traffic-readiness.