Network infrastructure can remain apparently available while inventories drift, undocumented links become critical, configurations diverge, support expires, capacity erodes, monitoring gaps grow, or redundant paths acquire shared dependencies. Individual projects install components; they rarely sustain the whole service over time.
The operating model assigns recurring responsibility for architecture, exceptions, inventory, topology, configuration, addresses, routing, capacity, security dependencies, telemetry, incidents, problems, changes, resilience exercises, support, spares, budgets, renewal, and retirement. It ensures that operational evidence can lead to an accountable decision rather than an endless handoff. This distinction also determines how capacity threshold and renewal roadmap should be evidenced and reconciled.