2.5.8 Observability & Monitoring

1. CloudWatch Logs

• Container logs from ECS tasks are sent to CloudWatch.

• Use log groups/filters to track errors, performance metrics.

2. CloudWatch Metrics & Alarms

• Monitor CPU, memory usage, request counts on ALB.

• Alarms can trigger scale-out actions or send notifications (via SNS) if usage spikes or if errors exceed thresholds.

3. Third-Party Monitoring (Datadog, New Relic, or similar)

• APM (Application Performance Monitoring) for deeper insight into Node.js performance, tracing requests from ALB → ECS → RDS.