Production Guides & Recipes

Lightweight Readiness Probing

To minimize database load during frequent health checks (e.g. every 5-10 seconds by Kubernetes):

import { payloadReadinessCheck } from "@nexload-sdk/healthcheck-payload";

const check = payloadReadinessCheck(payload, {
  collection: "users",
  // Excluded from liveness probe:
  scopes: ["readiness"],
  timeoutMs: 3000,
});

The check requests zero population depth (depth: 0) and limits results to 1 document, ensuring minimal memory and database CPU utilization.

---

Troubleshooting & Common Pitfalls

1. PAYLOAD_QUERY_FAILED

Indicates that the underlying database connection timed out or rejected the query. Check database pool exhaustion, connection string validity, or network partitions.

2. Readiness Probe Is Slow (>1 second)

Ensure the checked collection has an index and does not execute heavy afterRead hooks. If needed, create a dedicated minimal _health collection with a single static document.

3. Never Assign Payload Database Check to Liveness

If the database goes down, liveness failure will cause continuous pod restarts. Always assign database checks to scopes: ["readiness"].