EVPN Port-Active Redundancy Mode
Extracted elements (23)
For convergence in Port-Active mode, setting the standby port to LACP 'out-of-sync' (warm-standby) rather than operationally down improves transition times from standby to active. Prior synchronization of ARP/ND caches is recommended for IRB and L3 services; MAC table synchronization may be considered for L2 services.
Port-Active mode eliminates the need for ICCP and LDP, which are required by legacy MC-LAG solutions. This makes the solution agnostic to underlay technology (MPLS, VXLAN, SRv6) and enables efficient 1+N redundancy using BGP Route Reflectors, whereas ICCP requires a full mesh of LDP sessions.
Port-Active redundancy mode is introduced to provide deterministic traffic forwarding through a specific interface rather than statistical per-flow load balancing across multiple PEs. This determinism is essential for certain QoS features and ensures fast convergence during failure and recovery, which existing All-Active and Single-Active modes do not provide at the interface level.
Implementations complying with only RFC 7432 or RFC 8214 (predating this specification) will ignore the L2-Attr Extended Community in Ethernet A-D per ES routes and continue using default path resolution algorithms. The remote ESI Label Extended Community signals Single-Active redundancy mode, and P/B bits in Ethernet A-D per EVI routes are used as before.
A CE device connected to EVPN multihoming PEs MUST have a single LAG with all its links connected to the EVPN multihoming PEs in a redundancy group.
ES routes operating in Port-Active redundancy mode MUST signal the Single-Active multihomed site redundancy mode along with the Ethernet A-D per-ES route, and the ESI label-based split-horizon procedures of RFC 7432 Section 8.3 SHOULD be employed to prevent transient echo packets when L2 circuits are involved.
Identical LACP parameters MUST be configured on peering PEs, including the system ID, port priority, and port key.
Non-DF routers SHOULD implement a bidirectional blocking scheme for all traffic comparable to Single-Active redundancy mode, albeit across all VLANs. Non-DF routers MAY bring and keep the peering access interface in an operational down state, or MAY set the LACP state to Out of Sync (OOS) instead for better convergence.
P and B bits received in Ethernet A-D per EVI routes SHOULD be considered overridden by the 'parent' P and B bits when the L2-Attr Extended Community is advertised in the Ethernet A-D per ES route.
The AC-DF bit (A=1) MUST be set to 0 when advertising Port Mode DF Election capability (P=1). Upon receiving the AC-DF bit set (A=1) from a remote PE, it MUST be ignored when performing Port Mode DF Election, because AC (sub-interface) down events do not influence port-level DF election.
The DF router MUST keep the corresponding access interface in an up and forwarding active state for that ES.
The Ethernet Segment (ES) MUST be configured in Port-Active redundancy mode on peering PEs for the specified access interface.
The Ethernet Segment Identifier (ESI) MUST be assigned per access interface as described in RFC 7432; the ESI can be auto-derived or manually assigned and the access interface MAY be an L2 or L3 interface.
The primary/backup bits of the EVPN Layer 2 Attributes (L2-Attr) Extended Community (RFC 8214) SHOULD be used in Ethernet A-D per ES routes to achieve better convergence.
When the L2-Attr Extended Community is advertised in Ethernet A-D per ES routes, it SHALL have only the P or B bits set in the Control Flags field, and all other bits and fields MUST be zero. A remote PE receiving this community SHALL consider only the P and B bits and ignore other values.
A new capability bit (Bit 5, 'P' bit) is defined in the DF Election Capabilities bitmap of the DF Election Extended Community (RFC 8584): 'Port Mode Designated Forwarder Election'. When set, it signals that the DF election algorithm SHOULD be modified to consider the port ES only and not the Ethernet Tags.
For Port-Active mode, the modulo-based DF election algorithm operates at ES granularity only. The PE with ordinal i is the DF for an ES when (Es mod N) = i, where Es represents bytes 3-6 of the ESI and N is the redundancy group size. Bytes 3-6 are chosen because operators typically differentiate ESIs within these bytes.
Port-Active redundancy mode is a new EVPN multihoming mode where only one of the multihomed interfaces is in forwarding state at any time, with all services on the active interface operating in active mode and all services on the standby interface operating in standby mode. DF election operates per [ES] granularity rather than per [ES, Ethernet Tag].
The Highest Random Weight (HRW) algorithm for Port-Active mode omits the Ethernet Tag from the CRC computation; all references to (V, Es) are replaced by (Es) alone. DF and BDF are selected based on highest and second-highest Weight(Es, Si) values respectively, with ties broken by numerically lowest PE IP address.
When the Port Mode capability is signaled, preference-based DF election (RFC 9785) is modified to consider the port only and not associated Ethernet Tags. The Port Mode capability is compatible with the 'Don't Preempt' (D) bit, enabling non-revertive behavior at the port level upon interface recovery.
IANA has added Bit 5 ('Port Mode Designated Forwarder Election', Reference RFC 9786) to the 'DF Election Capabilities' registry under the 'Border Gateway Protocol (BGP) Extended Communities' registry group.
Introducing the Port Mode capability requires unanimity among PEs in a redundancy group. Without consensus, the DF election falls back to RFC 8584/RFC 7432 defaults. An attacker who modifies the configuration of one PE within an ES can exploit this fallback to force all PEs to revert to the default algorithm, potentially causing unfair load balancing, service disruption, traffic loss, or duplicate traffic.
The DF Election Capabilities bitmap (2 octets) in the DF Election Extended Community is amended to include Bit 5 (P) for Port Mode DF Election, alongside existing Bit 0 (D, Don't Preempt), Bit 1 (A, AC-DF), and Bit 3 (T, Time Synchronization).