Health Endpoint Monitoring

BeginnerFree3 min read

Expose health endpoints that external checks probe to detect failures.

Intent

Health Endpoint Monitoring has each service expose a /health endpoint that reports whether it and its key dependencies are working. Load balancers and monitors probe it to route around sick instances and alert on outages.

Diagram components: Health Endpoint Monitoring, Service.

Interactive diagram — open this article in RodGrid to explore it.

Probe the health endpoint

Diagram components: Coordinator, Health Endpoint Monitoring, Service.

Interactive diagram — open this article in RodGrid to explore it.

Tip Separate liveness (am I up?) from readiness (can I take traffic?) — they drive restarts vs load-balancer membership differently.

Common mistakes

Watch out Common mistake: merging liveness and readiness into one check, so a temporarily-busy-but-alive instance gets restarted instead of just pulled from the load balancer.

Try it on a Grid

Open a blank Grid and sketch a monitor probing a service's health endpoint, focusing on separating liveness from readiness.