DATABASES Signal 199
RFC 10018 defines BGP extensions for multicast and Ethernet VPNs over Segment Routing P2MP and Ingress Replication
↗
RFC 10018 specifies BGP encoding and procedure extensions for carrying multicast traffic over SR P2MP trees and Ingress Replication in BGP/MPLS IP VPNs and EVPNs within a Segment Routing domain.
This gives network operators a standards-track mechanism to build multicast provider tunnels using Segment Routing, covering both SR-MPLS and SRv6 data planes. It updates RFCs 6514 and 7988, so existing MVPN and IR implementations will need to account for the new encodings and procedures. PIM-based auto-discovery and Bidirectional PIM support are explicitly out of scope.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
The RFC extends BGP auto-discovery procedures for P-tunnels constructed with P2MP tree instances in SR domains.
It defines new SRv6 endpoint behaviors for MVPN with SR P2MP and IR P-tunnels, covering both SR-MPLS and SRv6 data planes.
It updates RFC 6514 and RFC 7988, and excludes PIM auto-discovery and Bidirectional PIM from its scope.
THE READ
What the cluster adds up to.
RFC 10018 fills a gap in how multicast traffic is tunneled across service provider networks that use Segment Routing. Previously, MVPN procedures in RFCs 6513 and 6514 defined P-tunnels for multicast, but the specific BGP encodings and procedures for instantiating those tunnels using SR P2MP trees or Ingress Replication were not standardized. This document provides those extensions for both BGP/MPLS IP VPNs and EVPNs.
The document covers two transport mechanisms: SR P2MP trees defined by SR P2MP Policy (RFC 9960) and Ingress Replication. An SR P2MP tree consists of a Root, a set of Leaf nodes, and candidate paths with optional constraints, instantiated via a controller such as a PCE. The RFC extends BGP auto-discovery to work with P2MP tree instances, and it extends the IR procedures from RFC 7988 to cover SR-MPLS with an SLA and introduces new procedures for SRv6 with or without an SLA.
For EVPN, the RFC addresses how P-tunnels for multi-destination traffic can be instantiated by SR-MPLS or SRv6 P2MP trees, building on the BGP MPLS-based EVPN framework in RFC 7432 as updated by RFC 9572. New SRv6 endpoint behaviors are defined specifically for MVPN with SR P2MP and IR P-tunnels, which means SRv6 implementations will need to add support for these behaviors to interoperate.
The scope boundaries are notable. PIM-based auto-discovery is excluded entirely, as is support for customer Bidirectional PIM. Operators relying on PIM for auto-discovery will not find coverage here and must look elsewhere. The document also defers scalability, operational, and troubleshooting considerations to RFCs 9960 and 9524, and OAM ping and traceroute procedures for SR P2MP Policy over SR-MPLS to RFC 9961.
Because this is a Proposed Standard on the Internet Standards Track, it represents IETF consensus and is intended for implementation. Vendors and operators maintaining MVPN or EVPN deployments that already follow RFC 6514 or RFC 7988 will need to track the updates this document makes to those procedures. The practical adoption cost depends on whether a deployment already uses Segment Routing; for those that do, this provides the missing standards-level glue for multicast.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER