Health Checks in Production APIs: Beyond Returning 200 OK

The /health endpoint that just returns { "status": "ok" } is useless. It tells you the process is running. It does not tell you if the app can actually serve requests.

Liveness vs Readiness

• Liveness: Is the process alive? If not, restart it.
• Readiness: Can it handle traffic? If not, stop sending requests.

app.get("/health/live", (_, res) => res.status(200).json({ status: "alive" }));

app.get("/health/ready", async (_, res) => {
  const checks = await Promise.allSettled([
    checkDatabase(),
    checkRedis(),
    checkExternalAPI(),
  ]);

  const results = {
    database: checks[0].status === "fulfilled",
    redis: checks[1].status === "fulfilled",
    externalAPI: checks[2].status === "fulfilled",
  };

  const healthy = Object.values(results).every(Boolean);
  res.status(healthy ? 200 : 503).json({ status: healthy ? "ready" : "not ready", checks: results });
});

Dependency Check Pattern

async function checkDatabase(): Promise<void> {
  const start = Date.now();
  await pool.query("SELECT 1");
  const latency = Date.now() - start;
  if (latency > 5000) throw new Error(`DB latency too high: ${latency}ms`);
}

async function checkRedis(): Promise<void> {
  await redis.ping();
}

Kubernetes Integration

livenessProbe:
  httpGet:
    path: /health/live
    port: 3000
  initialDelaySeconds: 5
  periodSeconds: 10
readinessProbe:
  httpGet:
    path: /health/ready
    port: 3000
  initialDelaySeconds: 10
  periodSeconds: 5

Do Not

• Do not check every downstream service in liveness. If Redis is down, restarting your app will not fix Redis.
• Do not let health checks take more than 2s total. Use timeouts.
• Do not require authentication on health endpoints. Load balancers need to reach them.

---

Part of my Production Backend Patterns series. Follow for more practical backend engineering.