SPRING

Meeting Minutes: IETF 125 SPRING Working Group Session

Summary

The SPRING Working Group met at IETF 125 to discuss the status of active drafts, review new proposals, and coordinate with other working groups, specifically IDR. The chairs emphasized the urgent need for increased working group participation and list engagement, noting that several critical documents (including security and YANG models) have seen low review activity. Key technical discussions focused on SRv6 Path Segment flag options, redundancy protection, inter-layer programming, and path verification.

Key Discussion Points

Administrivia and Document Status

RFC Editor Queue: draft-ietf-spring-sr-policy-yang and draft-ietf-dhc-srv6-ietf-dhc-srv6-locator-deployment have been approved.
IESG Processing: draft-ietf-spring-resource-aware-segments is in AD evaluation.
Security: draft-ietf-spring-srv6-security is undergoing an extended Last Call. Authors will update the draft based on SECINTOP and RTGWG directorate reviews before a subsequent WGLC.
Chair's Plea: Joel Halpern and Bruno Rijsman expressed disappointment regarding low engagement. Many adopted documents lack updates or list discussion, which delays consensus and impacts external dependencies (e.g., YANG).

Path Segment Identifier (PSID) in SRv6 (Segment Routing in IPv6)

Presenter: Guangming Zheng

Draft: draft-ietf-spring-srv6-path-segment
Discussion: The presentation analyzed six options for signaling the presence of a P-SID in the SRH (Segment Routing Header), categorized into dedicated flags, generic flags, and flagless mechanisms.
Technical Debate:
- Greg Mirsky questioned the necessity of P-SID in SRv6 compared to SR-MPLS and suggested using TLVs instead of 128-bit SIDs to save space.
- Cheng Li and Guangming Zheng noted that P-SID is needed for reduced mode or micro-SIDs where the full list is not available.
- Zafar Ali and Safar challenged the need for a special flag, suggesting that the endpoint can identify the P-SID via its own locator/function behavior without transit nodes needing to process it.
- Bruno Rijsman cautioned against "code point squatting," noting that implementations are using flags before IANA allocation or early allocation.

SRv6 for Redundancy Protection

Presenter: Balazs Varga

Draft: draft-ietf-spring-sr-redundancy-protection
Details: The draft defines a Redundancy SID (R-SID) for packet-level replication and elimination.
Discussion: The technical content is considered stable by the authors. Joel Halpern noted that while the content is stable, a WGLC cannot proceed without evidence of working group engagement on the mailing list.

SRv6 for Inter-Layer Network Programming

Presenter: Liuyan Han

Draft: draft-ietf-spring-srv6-inter-layer-programming
Details: Introduces the End.IL behavior for steering traffic into non-IP underlay links (Layer 1/Layer 2).
Hackathon Results: A multi-vendor project (Huawei, ZTE, FiberHome) demonstrated successful steering over MTN channels with stable latency and jitter even under congestion.

Linking BGP SR-TE to Spring concepts

Presenter: Susan Hares (IDR Chair)

Coordination: Susan Hares requested that SPRING authors provide clear links to SPRING architectural concepts in their BGP-related drafts. This is necessary for IDR to validate that protocol mechanisms align with the intended SR architecture.
Operator Input: Liuyan Han and Zafar Ali emphasized the importance of prioritizing features that are actually deployed in operator networks.

SRv6 Path Verification

Presenter: Fan Yang

Proposal: A recursive authentication algorithm at each hop to ensure the packet strictly follows the SID list, targeting enterprise security scenarios.
Discussion: Greg Mirsky suggested reviewing previous "Proof of Transit" work in the SFC (Service Function Chaining) WG. Zheng Sun recommended adding capability negotiation for backwards compatibility.

SID as source address in SRv6

Presenter: Fan Yang

Proposal: Using the Service SID as the IPv6 Source Address to make SRv6 traffic "stateful firewall friendly" by ensuring symmetric address session tables.
Discussion: Zafar Ali raised concerns regarding ICMP error handling when the source address is a VPN SID. Bruno Rijsman suggested presenting this to the BESS WG and broadening the scope beyond VPN use cases.

4map6 Segments for IPv4 Service delivery over IPv6-only underlay networks

Presenter: Guozhen

Proposal: A new SID type to support IPv4-IPv6 conversion based on mapping rules in multi-domain IPv6-only underlays.
Next Steps: The authors requested working group adoption.

Encapsulation of BFD for SRv6 Policy

Presenter: Xiaoming

Proposal: Defines "Insert" and "Encap" modes for BFD for SRv6 Policy.
Discussion: Joel Halpern cautioned that "Insert mode" must only be performed by the node originating the packet to avoid RFC 8200 violations. Ketan Talaulikar and chairs suggested the draft might be better suited for the srv6ops working group if it describes existing implementation practices.

Decisions and Action Items

Draft Split: Authors of draft-ietf-spring-sr-service-programming have requested to split the document into MPLS and SRv6 specific drafts. Chairs will initiate a list discussion.
Security Draft: Authors of draft-ietf-spring-srv6-security to update the document based on directorate reviews before the chairs re-issue a WGLC.
Coordination: SPRING chairs to coordinate with srv6ops chairs regarding the appropriate home for BFD encapsulation documentation.

Next Steps

Participation: All members are urged to review active drafts, particularly those nearing WGLC or with external dependencies.
P-SID: Discussion on flag vs. flagless options for draft-ietf-spring-srv6-path-segment to continue on the mailing list.
Inter-layer Programming: Authors to continue control plane extensions in related working groups (e.g., PCE/IDR/LSR).

Automatic IETF Minutes