Network Working Group P. Mykhailyk Internet-Draft 16 November 2026 Intended status: Standards Track Expires: May 16, 2027 BGP Extended Community Exchange draft-pmykhailyk-idr-bgp-community-exchange-00 Status of This Memo This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at https://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." This Internet-Draft will expire on May 16, 2027. Copyright Notice Copyright (c) 2026 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Abstract This document specifies a mechanism for transferring a BGP Extended Community, in particular the Link Bandwidth Extended Community, from one prefix to another, which enables routers to dynamically adjust the distribution ratio for weighted load balancing in NAT scenarios, allowing a router to perform unequal cost load balancing according to the real capacity of compute clusters. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1. Requirements Language . . . . . . . . . . . . . . . . . . . 2 2. BGP Extended Community Exchange Mechanism . . . . . . . . . . . 2 P. Mykhailyk Expires 16 May 2027 [Page 1] Internet-Draft BGP Extended Community Exchange 16 November 2026 3. Deployment Considerations . . . . . . . . . . . . . . . . . . . 3 4. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . 3 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 3 6. Security Considerations . . . . . . . . . . . . . . . . . . . . 3 7. References . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7.1. Normative References . . . . . . . . . . . . . . . . . . . 3 7.2. Informative References . . . . . . . . . . . . . . . . . . 4 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 4 1. Introduction Load balancing is a critical aspect of network design, enabling efficient utilization of available bandwidth and improving overall network performance. In BGP [RFC4271], the Link Bandwidth Extended Community (LBEC) [RFC10005] provides a way to perform WECMP (Weighted ECMP) and adopt a distribution ratio to metrics encoded in prefix attributes. In many cases, however, the real capacity metric for a compute cluster can be provided only by the cluster management system or by some sort of orchestration engine. The main problem is that in most cases the network layer that performs load balancing between DCs operates with public VIPs of services [RFC4786], while the cluster orchestrator announces private VIPs, and one or more NAT translations occur on the traffic path from the network load balancing level to the ingress point of the compute cluster. In this situation, it is impossible to directly use capacity metrics encoded in the BGP Link Bandwidth Extended Community at the network DCI level. This document describes a proposal to add the possibility to transfer Extended Community values from one prefix to another via a route-map extension. In this case, capacity metrics can be directly injected into the network layer even for NAT translation scenarios, and the distribution ratio can be dynamically adjusted to the real capacity of the cluster. 1.1. Requirements Language The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here. 2. BGP Extended Community Exchange Mechanism The BGP Extended Community exchange is defined as an extension to the route-map mechanics that allows adding rules to match a target prefix and transfer a BGP Extended Community [RFC4360] carrying the bandwidth information to a locally generated route. This community can be used to inform other routers about the available capacity of the current compute cluster. The transfer mechanism MUST allow matching the Type and Subtype values [RFC4360] of the communities that will be transferred. P. Mykhailyk Expires 16 May 2027 [Page 2] Internet-Draft BGP Extended Community Exchange 16 November 2026 3. Deployment Considerations Originator - a BGP speaker that originates a Link Bandwidth Extended Community [RFC10005]. A BGP speaker that attaches a Link Bandwidth Extended Community corresponding to the actual capacity of the current compute cluster SHOULD be able to re-advertise the route with an updated LBEC when the cluster capacity changes during autoscaling or failure scenarios. Translator - a device that provides LBEC transfer. The Translator MUST be able to match a 'source' route that carries an LBEC and transfer the Link Bandwidth Extended Community from the 'source' route to a Route Health Injection route (RHI) that is normally generated for the eVIP according to the service health-check status. NLB - a device that performs WECMP. The NLB MUST be able to process the LBEC and use it to provide WECMP according to the metrics encoded in the LBEC. In case the device that originated the RHI is unable to provide LBEC transfer, the NLB can be used in the Translator role and provide LBEC transfer for the received routes. The Extended Route-Map Mechanics (ERMM) MUST allow setting an 'exact match' condition for the 'source route' with the Type and Subtype of the community that will be used for the SET operation on the target route. ERMM SHOULD be available for advertised routes on the Translator, or for received routes on the NLB. 4. Acknowledgments The author would like to thank Pradosh Mohapatra and Rex Fernando for introducing the idea of the BGP Link Bandwidth Extended Community. 5. IANA Considerations This document does not require any actions from IANA. It defines an extension to the route-map mechanics to allow direct exchange of compute cluster capacity metrics via the Link Bandwidth Extended Community in NAT scenarios. 6. Security Considerations There are no additional security risks introduced by this design. 7. References 7.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997, . P. Mykhailyk Expires 16 May 2027 [Page 3] Internet-Draft BGP Extended Community Exchange 16 November 2026 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, May 2017, . 7.2. Informative References [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and Hares, S., Ed., "A Border Gateway Protocol 4 (BGP-4)", RFC 4271, January 2006, . [RFC4360] Sangli, S., Tappan, D., and Rekhter, Y., "BGP Extended Communities Attribute", RFC 4360, February 2006, . [RFC4786] Abley, J., and Lindqvist, K., "Operation of Anycast Services", BCP 126, RFC 4786, December 2006, . [RFC10005] Mohapatra, P., Das, R., Ed., Mohanty, S., Ed., Krier, S., Szarecki, R.J., and Gattani, A., "BGP Link Bandwidth Extended Community", RFC 10005, June 2026, . Authors' Addresses Pavlo Mykhailyk Email: pavel.mykhailyk@gmail.com P. Mykhailyk Expires 16 May 2027 [Page 4]