Routing Protocols

BGP multipath prioritization with configurable priority queues (PTX10001-36MR and QFX5130-48C)—Use BGP multipath prioritization to prioritize critical routes in BGP multipath computations with queues of low, medium, and high priorities. This feature reduces multipath-calculation latency for operator-prioritized routes in overall route convergence during high load. Additionally, you can manage delayed forwarding information base (FIB) convergence that results from frequent routing changes.

[See BGP Route Prioritization and multipath (Protocols BGP).]

BGP route leak prevention by using BGP roles and OTC attributes based on RFC 9234 (ACX7100-32C, ACX7100-48L, ACX7332, ACX7348, ACX7509, ACX7024, ACX7024X, PTX10001-36MR, PTX10002-36QDD, PTX10003, PTX10004, PTX10008, PTX10016, QFX5130-32CD, QFX5130E-32CD, QFX5130-48C, QFX5130-48CM, QFX5140-24CD8O, QFX5700, QFX5700E, QFX5220, QFX5230-64CD, QFX5240-64OD, QFX5240-64QD, QFX5241-32OD, QFX5241-32QD, QFX5241-64OD, QFX5241-64QD, QFX5241E-64OD, QFX5250-64OE, and QFX5250-64OE-DOT2L)—You can prevent route leaks in BGP routing by using BGP roles and OTC attributes as defined in RFC 9234. The feature ensures that routes from providers or peers are only propagated to customers, reducing misconfigurations and errors. The BGP speaker automatically sets the OTC based on its configured role, and then advertises a prefix based on the OTC presence in the BGP update message, making the configuration straightforward and minimizing manual intervention. With this feature, you can maintain intended routing policies and prevent network delays and denial-of-service (DoS) attacks.

Enhanced service route resolution for BGP multipath with list next hop (ACX7100-32C, ACX7100-48L, ACX7332, ACX7348, ACX7509, ACX7024, ACX7024X, PTX10001-36MR, PTX10002-36QDD, PTX10003, PTX10004, PTX10008, PTX10016, PTX12008, QFX5130-32CD, QFX5130E-32CD, QFX5130-48C, QFX5130-48CM, QFX5700, QFX5700E, QFX5220, QFX5230-64CD, QFX5240-64OD, QFX5240-64QD, QFX5241-32OD, QFX5241-32QD)—This feature improves how service routes resolve over BGP multipath routes that use list-next hop structures. The resolver now tracks all contributing paths, not just the active one. When a service route resolves to BGP multipath helper route, RPD builds internal dependencies across each indirect next hop in the list. This ensures that your service route automatically re-resolves whenever any contributing path changes.

Previously, inactive paths could change without triggering a re-resolution. To fix this issue, RPD now links multipath routes to their full set of contributing routes using a patricia tree structure. The resolver can now detect changes across the entire path set and update service routes as needed.

This enhancement improves resolution accuracy, supports re-evaluation during inactive path changes, and strengthens overall routing consistency in hybrid internal BGP (IBGP) and external BGP (EBGP) environments.

[See Understanding BGP Path Selection.]