**Session Date/Time:** 10 Aug 2023 14:00

# [MPLS](../wg/mpls.html)

## Summary

This interim meeting focused primarily on the interaction between MPLS Network Actions (MNA) and Bit-Indexed Explicit Replication (BEER) in an MPLS network, particularly concerning Post-Stack Data (PSD) MNA. The discussion highlighted the architectural challenges posed by BEER's current encapsulation (expecting its label as the bottom of the stack) and explored several potential solutions, none of which were without significant implications for existing deployments or RFCs. Updates were also provided on the status of the MNA requirements and framework drafts, as well as the MPLS MNA header solution draft.

## Key Discussion Points

*   **Administrative and Agenda Review**
    *   The chairs welcomed attendees and presented the IETF Note Well.
    *   The agenda included a review of outstanding action items, discussion on BEER with MNA, clarification questions for the MNA header solution draft authors, and future agenda items.
    *   The third agenda item (MNA header solution draft) was clarified as a chance for authors to ask clarification questions, not a full discussion.

*   **Review of Outstanding Action Items**
    *   **Requirements/Framework Draft Merge**: The possibility of merging the MNA requirements and framework drafts, initially raised by Matthew and to be followed up by Tony, had not progressed.
        *   Eric indicated that he is not in favor of merging the documents, a sentiment he felt was shared by others.
        *   The requirements draft has passed routing director review and is ready to progress, pending a final discussion with Matthew.
    *   **MPLS MNA Header Draft Updates**:
        *   Updates were discussed regarding suggestions made on the mailing list and during IETF 117 sessions, including text from Joel.
        *   Authors were tasked with providing feedback on these suggestions.
        *   A fuller discussion on the MNA header document is anticipated at the next meeting.

*   **BEER with MNA in MPLS Networks**
    *   **Overview of BEER over MPLS**: Greg presented on BEER, describing it as a multicast technology that works in the data plane without per-flow state in transit nodes. The BEER header includes a Bit Index Forwarding (BIF) ID, which acts as a BEER MPLS label and is currently expected to be the bottom-of-stack (BoS) label.
    *   **Interaction with MNA**:
        *   **Instack Data (ISD) MNA**: It was generally agreed that ISD MNA is largely transparent to BEER, as it operates within the MPLS label stack and does not interfere with the BEER header itself. However, G raised a point about propagating ISD information across multiple BEER nodes if the label stack is popped and rebuilt. Jeffrey clarified that MPLS is merely a transport between BEER routers, and MPLS information (including MNA) typically terminates at each BEER router.
        *   **Post-Stack Data (PSD) MNA**: The primary challenge for MNA with BEER arises with PSD MNA, due to BEER's expectation that its MPLS label is the bottom of the stack, immediately followed by the rest of the BEER header. This prevents the insertion of PSD MNA directly after the BEER MPLS label.
    *   **Proposed Solutions for PSD MNA and BEER**: Three options were discussed for integrating PSD MNA with BEER:
        1.  **Present PSD MNA as part of the BEER payload**: Define a new BEER protocol ID for a PSD MNA block to follow the BEER header. This would place MNA at the BEER layer, with replication across the BEER domain. Concerns were raised that this makes MNA specific to BEER, rather than a generic MPLS functionality.
        2.  **Not setting the S-bit in the BEER MPLS LSE**: This would allow further MPLS labels (including MNA labels) to follow the BEER MPLS label. However, Tony P and Loa emphasized that this directly violates current BEER RFCs and would break existing implementations, making it a very costly option.
        3.  **Architecturally separate BEER MPLS label from the rest of the BEER header**: This approach, suggested by Jeffrey, would allow PSD MNA to be inserted between the BEER MPLS label and the subsequent BEER header elements. This option requires careful consideration of backward compatibility with existing BEER routers.
    *   **Lack of Concrete Use Case**: It was noted that a concrete use case for PSD MNA specifically with BEER has not yet been clearly identified, making the discussion largely theoretical at this stage.

*   **MPLS MNA Header Solution Draft Clarifications**
    *   The authors of the MPLS MNA header solution draft did not have any clarification questions for the working group.

*   **Future Agenda Items**
    *   A reminder was given that items are collected for future interims and that interims may be canceled if there are no pressing discussion points.

## Decisions and Action Items

*   **Requirements/Framework Draft**: Eric to coordinate with Matthew on the progression of the requirements draft. The proposal to merge the requirements and framework drafts is unlikely to proceed.
*   **MPLS MNA Header Draft**: Authors are to provide feedback on the proposed updates (Joel's text and other mailing list comments) to the working group. A discussion on this draft is targeted for the next meeting.
*   **BEER with MNA Discussion**: The discussion on BEER and MNA interactions, particularly the proposed solutions for PSD MNA, will continue on the mailing list, leveraging Jeffrey's recent email thread. The chairs will review the feedback and organize follow-up.

## Next Steps

*   Chairs to follow up on the BEER with MNA discussion, potentially involving the BEER WG as well, given the architectural implications.
*   Authors of the MPLS MNA header solution draft to provide their responses to the suggested updates.
*   Working group participants are encouraged to propose items for future interim meetings to ensure continued progress.