How do PE routers exchange CE routes between remote sites?

Question

Options

• Aby converting CE routes into VPNv4 routes and exchanging them using MP-BGP
• Bby establishing BGP neighbor relationships between all connected CEs to exchange routing
• Cby learning IPv4 routes from connected CEs and redistributing them into the global IGP
• Dby converting CE routes into VPNv4 routes and exchanging them using the global IGP

• A
  75% (12)
• B
  6% (1)
• C
  6% (1)
• D
  13% (2)

Explanation

Option A is correct because MPLS L3VPN relies on PE routers importing CE routes into a VRF, then prepending a Route Distinguisher (RD) to make them globally unique VPNv4 addresses. MP-BGP (Multiprotocol BGP) is the only protocol capable of carrying these VPNv4 NLRI across the provider core while also transporting Route Target (RT) extended communities used for import/export policy.

B is wrong because CE routers only peer with their local PE - they have no direct neighbor relationship with remote CEs. Full-mesh CE-to-CE BGP would destroy the scalability and isolation that MPLS VPN exists to provide.

C is wrong because customer routes are kept isolated inside per-customer VRFs, never leaked into the provider's global IGP. The global IGP (e.g., OSPF/IS-IS) carries only provider infrastructure prefixes (loopbacks, P-to-PE links) needed for LSP forwarding.

D is wrong because the global IGP has no mechanism to carry VPNv4 routes or attach RT communities. Only MP-BGP supports the VPNv4 address family required for this function.

Memory tip: MP-BGP = "Multi-Protocol" - its whole value is carrying more than just IPv4, including VPNv4. Whenever a question involves cross-site VPN route exchange on a PE, MP-BGP is the answer.

No community discussion yet for this question.