Summary
- A service map should show physical paths, routing, dependencies and accountable operators together.
- Redundancy counts only when power, fibre, facilities, software and response teams do not share one failure.
Packets move because many independent networks agree on interfaces and exchange reachability, not because a central backbone carries everything. DNS finds services, routing selects paths, cables and radio carry signals, and data centres run applications. The next useful evidence is an end-to-end failure drill that removes one dependency and records user impact, detection and recovery. Infrastructure becomes understandable when each layer has an owner and tested boundary, rather than disappearing behind a cloud-shaped diagram.
