pattern·pro·health_endpoint
/health endpoint
Liveness probe that monitoring + load balancers hit. Returns 200 + version + status.
Two flavors: liveness (process is up) vs readiness (process can serve real traffic — DB roundtrip, dependencies). Don't make /health do a DB query if your monitoring is hammering it 1/sec — but a "no DB roundtrip" /health is a 25-hour-down-without-noticing waiting to happen. Tradeoff matters.
Beziehungen
opspatternmonitoring