IDR Working Group W. Cheng Internet-Draft China Mobile Intended status: Standards Track C. Lin Expires: 1 May 2027 New H3C Technologies 28 October 2026 BGP Extension for ECMP Group ID draft-cheng-idr-ecmp-groupid-00 Abstract In AI training scenarios, per-flow ECMP is the mainstream load balancing technology. However, hash collision issues lead to imbalanced traffic distribution, adversely affecting application performance. This document defines a new BGP Extended Community, BGP_EXT-COMM- ECMP-GROUPID, which carries an ECMP Group ID associated with a route. The ECMP Group ID allows an ingress router to map flows destined for different remote addresses to different local ECMP subgroups, providing finer-grained load distribution and resolving load imbalance caused by hash collisions. 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 1 May 2027. Copyright Notice Copyright (c) 2026 IETF Trust and the persons identified as the document authors. All rights reserved. Cheng & Lin Expires 1 May 2027 [Page 1] Internet-Draft BGP ECMP GroupID October 2026 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. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 3 2. Protocol Extension . . . . . . . . . . . . . . . . . . . . . 3 3. Security Considerations . . . . . . . . . . . . . . . . . . . 4 4. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 4 5. References . . . . . . . . . . . . . . . . . . . . . . . . . 4 5.1. Normative References . . . . . . . . . . . . . . . . . . 4 5.2. Informational References . . . . . . . . . . . . . . . . 4 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 4 1. Introduction Currently, there are two granularities for network load balancing: per-flow ECMP and per-packet forwarding. The per-flow ECMP method employs flow characteristic-based hashing (typically using the five-tuple) to distribute traffic across multiple ECMP paths. Its primary advantage is the elimination of packet reordering issues. However, this method presents limitations when dealing with either a limited number of flows, or the presence of elephant flows. In such cases, five-tuple-based hashing may lead to hash collisions, causing disproportionate mapping of oversized flows to the same path. This results in suboptimal load balancing performance. The per-packet forwarding approach applies hashing to each individual packet. Theoretically, it achieves optimal load-balancing granularity. However, it introduces severe packet reordering within the same flow, necessitating additional mechanisms (e.g., reordering buffers or sequence tracking) to handle out-of-order delivery. In AI training scenarios, the current mainstream load balancing technology is per-flow ECMP. However, hash collision issues lead to imbalanced traffic distribution, adversely affecting application performance. Cheng & Lin Expires 1 May 2027 [Page 2] Internet-Draft BGP ECMP GroupID October 2026 To address this problem, an enhanced ECMP method that resolves load imbalance caused by hash collisions has been proposed in [I-D.cheng-rtgwg-enhanced-ecmp]. In one of the proposed methods, the ECMP egress interfaces on the ingress router are divided into subgroups, and flows are mapped to subgroups based on the remote egress group index carried in the route. This requires the route to carry such an egress group index. This document defines a new BGP Extended Community to carry the ECMP Group ID associated with a route, as required by the enhanced ECMP mechanism. 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. Protocol Extension This document defines a new extended community attribute type [RFC4360] to carry the ECMP ID associated with a BGP route [RFC4271]. The ID comprises a 4-byte Router ID and a 2-byte Group-ID. The format is as follows: 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | 0x83 | Sub-Type(TBD) | Router ID | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Router ID | Group-ID | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 1: ECMP GroupID Extended Community Sub-Type: BGP_EXT-COMM-ECMP-GROUPID (TBD) Value Structure: * Local BGP RouterID (4 bytes) * ECMP-Group-ID Value (2 bytes) When a route is advertised with the enhanced ECMP mechanism described in [I-D.cheng-rtgwg-enhanced-ecmp], the route MUST carry this extended community. The Router ID field MUST be set to the local BGP Cheng & Lin Expires 1 May 2027 [Page 3] Internet-Draft BGP ECMP GroupID October 2026 RouterID of the advertising device, and the ECMP-Group-ID field MUST be set to the egress group index assigned on the advertising device. A router that receives this extended community but does not support the enhanced ECMP mechanism MAY ignore it and perform normal ECMP forwarding. 3. Security Considerations TBD. 4. IANA Considerations Registry Name: Transitive BGP_EXT-COMM-ECMP-GROUPID Community Sub- Types TBD: BGP_EXT-COMM-ECMP-GROUPID 5. References 5.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997, . [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway Protocol 4 (BGP-4)", RFC 4271, DOI 10.17487/RFC4271, January 2006, . [RFC4360] Sangli, S., Tappan, D., and Y. Rekhter, "BGP Extended Communities Attribute", RFC 4360, DOI 10.17487/RFC4360, February 2006, . [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, May 2017, . 5.2. Informational References [I-D.cheng-rtgwg-enhanced-ecmp] Cheng, W. and C. Lin, "Enhanced ECMP for AI Cluster", Work in Progress, Internet-Draft, draft-cheng-rtgwg-enhanced- ecmp-01, 25 March 2026, . Authors' Addresses Cheng & Lin Expires 1 May 2027 [Page 4] Internet-Draft BGP ECMP GroupID October 2026 Weiqiang Cheng China Mobile Beijing China Email: chengweiqiang@chinamobile.com Changwang Lin New H3C Technologies Beijing China Email: linchangwang.04414@h3c.com Cheng & Lin Expires 1 May 2027 [Page 5]