Visualize BGP with Containerlab and FRRouting Dashboard

Make BGP Visible: A Live Topology Dashboard with Containerlab

Gergo Vadasz developed a custom BGP dashboard using Python and Containerlab to move away from heavyweight GUI-based simulators. The system manages a four-router topology with a memory footprint of only 350 MB while providing real-time WebSocket updates.

Why This Matters

Traditional network simulators like GNS3 or EVE-NG often require significant hardware resources and proprietary images, creating barriers to entry for rapid prototyping. By utilizing FRRouting in Docker containers, engineers can implement Infrastructure-as-Code (IaC) workflows where network topologies are version-controlled and deployed in seconds.

Key Insights

• FRRouting (FRR) provides production-grade, open-source routing software that avoids licensing hurdles common in vendor-specific simulators.
• Containerlab uses a single YAML file to define network topology, enabling lab environments to be managed as Git-tracked assets.
• Real-time monitoring is achieved by polling vtysh for JSON data every two seconds and pushing diffed state changes via WebSockets.
• The dashboard visualizes BGP Finite State Machine (FSM) transitions and Routing Information Base (RIB) updates instantly without manual CLI polling.
• Small-scale topologies of two ISPs and two customers allow for testing complex concepts like AS_PATH prepending and LOCAL_PREF in a simplified environment.

Working Examples

Commands to clone and deploy the BGP lab environment using Containerlab.

git clone https://github.com/vadaszgergo/bgp-lab-with-dashboard
cd bgp-lab-with-dashboard
sudo containerlab deploy -t simple.clab.yaml

Practical Applications

• Use case: Simulating multi-homing and failover scenarios between customer networks and multiple ISPs to observe real-time reconvergence. Pitfall: Running the dashboard with an exposed Docker socket without authentication in non-trusted environments.
• Use case: Validating BGP community tagging and local preference policies in an Infrastructure-as-Code pipeline. Pitfall: Relying on manual CLI output scrolling instead of visual state diffing, which can obscure transient routing changes.

References:

• https://github.com/vadaszgergo/bgp-lab-with-dashboard
• https://www.youtube.com/watch?v=N1RKT6id1UI
• https://dev.to/gergovadasz/make-bgp-visible-a-live-topology-dashboard-with-containerlab-3a54