Set Up Monitoring

When setting up monitoring for a service:

1. Add a health check endpoint that verifies the service and its critical dependencies are reachable.
2. Add structured logging with consistent fields: timestamp, level, message, request_id.
3. Add metrics for: request count, error rate, response time (p50, p95, p99), and active connections.
4. Set up dashboards showing these metrics over time.
5. Configure alerts for: error rate exceeding baseline by 2x, p99 latency exceeding SLA, health check failures.
6. Ensure each alert has a runbook linked in the alert message.
7. Test the alerting pipeline: trigger a test alert and verify it reaches the on-call channel.