“Full-mesh iBGP requires N*(N-1)/2 peerings. In a core network with 20 routers, that’s 190 iBGP neighbor sessions. Deploying Route Reflectors reduces this to 20 hub sessions.”

While managing Net4 India’s ISP backbone as Assistant Manager, scaling internal BGP (iBGP) across 15 datacenter routers was consuming router CPU and memory due to full-mesh peering constraints.

Architecture: BGP Route Reflectors & Inbound Traffic Engineering

We re-architected the ISP core network using BGP Route Reflectors (RR) and outbound AS-Path Prepending.

1. Route Reflector Hubs: Configured dual Cisco ASR 9006 core routers as iBGP Route Reflectors, allowing client routers to receive BGP updates without maintaining full-mesh peerings.
2. Inbound Traffic Steering (AS-Path Prepending): Steered inbound customer traffic away from congested secondary ISP links by prepending our AS number (AS 17813 AS 17813 AS 17813) on backup transit BGP sessions.

[!TIP] Always deploy dual Route Reflectors in separate physical racks for high availability, configuring identical cluster-id 1.1.1.1 to prevent routing loops.

# # Cisco Router BGP Route Reflector & AS-Path Prepend Configuration
router bgp 17813
 bgp cluster-id 1.1.1.1
 neighbor 192.168.1.10 remote-as 17813
 neighbor 192.168.1.10 route-reflector-client

route-map PREPEND-ISP2-OUT permit 10
 match ip address prefix-list PUBLIC-PREFIXES
 set as-path prepend 17813 17813 17813

The Verdict

Key Takeaway

Deploy BGP Route Reflectors to Scale Internal Core Networks Beyond Full-Mesh Limits.

Implementing BGP Route Reflectors and AS-Path Prepending simplifies core router configuration while delivering deterministic inbound traffic steering across ISP links.