Node Watch v1.0.0 — Initial Release

Initial release of Node Watch, a lightweight node-level telemetry agent designed to demonstrate real-world infrastructure monitoring patterns.

Key capabilities:

• Host vs runtime telemetry separation
• System metrics collection (CPU, memory, disk, uptime)
• Prometheus-compatible metrics endpoint (/metrics)
• HTTP API for structured telemetry (/node)
• Containerized runtime with Docker
• Kubernetes deployment via DaemonSet (one agent per node)
• AWS deployment using ECR and EC2
• External probe integration using Go

This project is designed as a practical infrastructure engineering exercise, focusing on how monitoring agents behave across environments rather than building a full monitoring platform.

Validated environments:

• Local Python runtime
• Docker container runtime
• Kubernetes (kind cluster)
• AWS EC2 with ECR-backed image

Notes:

• Host-level telemetry requires explicit filesystem mount (/host)
• Telemetry gracefully degrades when host data is unavailable

Future work:

• Expanded telemetry coverage (network, temperature)
• Alerting and threshold evaluation
• Multi-node aggregation