**Session Date/Time:** 23 Jul 2024 00:30

# mpls

## Summary

This MPLS working group meeting covered several key topics including status updates on existing documents, discussion of Substack and Post-Stack solutions, IOWAN/IOWAM considerations, and deterministic latency concerns. Key discussions revolved around the MNA header format, the utility of Post-Stack solutions for IOM, and the viability of deterministic latency mechanisms. Several outstanding questions remain regarding the practicality and performance implications of different approaches, particularly concerning RLD limitations.

## Key Discussion Points

*   **MNA Header Update (Substack):** Updates were presented regarding the MNA header, particularly concerning format B and C, and the inclusion of a U-bit for unknown network action handling.
*   **Post-Stack MNA:** Discussions centered on the Post-Stack MNA approach, particularly for IOM use cases.  Greg Mirsky questioned the necessity of Post-Stack solutions when ISD might suffice.  Joel Halpern raised concerns about how Post-Stack MNA would function when the label stack exceeds the readable label depth (RLD). The pros (ECMP compliance) and cons (performance) of moving IOM specific actions and data to the post-stack were discussed.
*   **IOM/IOWAM:** Greg Mirsky raised a concern regarding the necessity of post-stack data specifically for IOM direct export.  Rakesh Gandhi argued for using post-stack data for various IOM option types, citing advantages for ECMP hashing, extensibility and RLD concerns, even though data is moved to post-stack.
*   **Deterministic Latency:** Rakesh Gandhi presented on deterministic latency, explaining that the goal is measurable and controllable latency. There was a discussion about presenting the work to the DETNET working group for feedback.

## Decisions and Action Items

*   **MNA Header Update:** Authors were asked to confirm that all outstanding issues raised during the first working group last call have been addressed or send a reminder email to commenters.
*   **Post-Stack MNA:** Authors to consider the efficiency of ancillary data encoding if the data is to be used by multiple actions.
*   **Deterministic Latency:** Present the draft in the DETNET working group to get their feedback.

## Next Steps

*   Address outstanding comments and issues related to the MNA header draft.
*   Continue discussions regarding the Post-Stack MNA approach on the mailing list, focusing on use cases and RLD concerns.
*   Seek working group adoption of the Post-Stack MNA draft.
*   Present deterministic latency solution to the DETNET working group.