BGP Auto-Discovery (AD) for LDP VPLS

Overview

The BGP Auto-Discovery enables automatic discovery of VPLS peers, eliminating the need for manual peer configuration. Once discovered, pseudo-wires (PWs) between peers are established using LDP signaling, streamlining the VPLS setup process.

OcNOS supports both BGP Auto-Discovery (BGP-AD) and manual VPLS peer configuration within the same VPLS instance. This is beneficial in topologies where certain VPLS peers do not support BGP-AD, as they can still establish VPLS sessions using manual peer configuration.

In BGP AD VPLS, if rd and/or router-target is configured in ASN4:nn format while BGP extended ASN support is disabled or vice-versa then BGP auto-discovery will not work and mesh PW will not come up. In such scenarios user is expected to remove the BGP AD VPLS instance and configure it again with correct rd and route-target config as per the BGP extended ASN support.

Benefits

The BGP Auto-Discovery provides in following aspects:

•

Simplifies the VPLS configuration process.

•

Enhances network scalability.

•

Improves scaling efficiency when used with route reflectors.

Limitations

•

MIB is not supported.

•

The PW group feature is not supported in OcNOS for both FEC128 and FEC129 signaling in BGP AD VPLS. During interoperability, avoid configuring PW groups on peer devices, as related wildcard label messages will be ignored. However, OcNOS now parses PW Group ID TLVs with a group ID of 0— commonly sent by some vendors—even if PW groups are not configured, ensuring interoperability.

•

For using a local IP address for BGP Auto-Discovery (AD) NLRI advertisement and as the SAII in FEC129-based pseudowire (PW) establishment, the system by default selects the secondary IP address on the lo (loopback) interface. If no secondary IP is configured, it falls back to the BGP router ID. However, if a secondary IP is configured on the loopback interface and it differs from the BGP router ID, BGP AD will fail to operate correctly. Additionally, the LDP transport address must match the selected local IP address (i.e., the secondary loopback IP or router ID) to ensure consistency across control-plane protocols.