Automating Network Assurance with NetBox Labs and Cisco Catalyst Center

This Lab (Managing Device Data Models for Use as Automation Dynamic Inventory) has three modules that build a closed-loop network assurance solution utilizing Cisco Catalyst Center, NetBox, and Ansible.

Challenge 1: establishes the source-of-truth pipeline: Orb Agent (NetBox Labs) polls Catalyst Center every 15 minutes and forwards device/site data via gRPC to Diode, which idempotently writes it into NetBox — including custom fields like catalyst_device_id and catalyst_serial_number. The lab has you verify the agent's systemd/Docker health and confirm synced data in the NetBox UI.

Challenge 2: turns that source of truth into live automation: after discovering devices in Catalyst Center, you generate a NetBox API token, wire it into the NetBox dynamic inventory plugin, and use ansible-inventory to query devices by custom field (cf_catalyst_device_id). You then run a playbook that uses NetBox-sourced device IDs to pull live uptime from Catalyst Center first for a core switch, then an edge router, showing how inventory metadata drives targeted automation.

Challenge 3: closes the loop with operational intelligence: a playbook queries Catalyst Center's device-health API (overallHealth score, 1–10), evaluates it against a threshold, and writes the result back into NetBox as a custom field (assurance_health_status: Healthy/Degraded). This gives downstream tools live health context, not just static inventory, and the lab notes that a production version of this would be event-driven (via Ansible EDA) rather than run on demand.

Outcome: Catalyst Center → Orb Agent/Diode → NetBox keeps inventory current automatically, Ansible + NetBox dynamic inventory turns that data into actionable automation, and the health-score playbook extends NetBox from a static CMDB into a live assurance-aware source of truth.

1. Maintain an accurate, always-current source of truth by automatically syncing device inventory from Cisco Catalyst Center into NetBox (via Orb Agent and Diode), eliminating manual imports and inventory drift.
2. Enable dynamic, source-of-truth-driven automation by using NetBox as a live Ansible inventory — letting playbooks target devices through custom-field filters instead of static host lists.
3. Extend inventory into operational intelligence by capturing live device health scores from Catalyst Center and writing them back into NetBox, so automation can act on current device state, not just static attributes.
4. Lay the groundwork for event-driven assurance, where health degradations trigger remediation or alerting automatically (e.g., via Ansible EDA) rather than relying on scheduled polling or manual dashboard checks.