**Session Date/Time:** 21 Mar 2024 03:00

# lsr

## Summary

The LSR working group session covered a range of topics, including working group document status, updates on ongoing drafts, and presentations of new proposals. Key discussions revolved around the adoption of drafts related to multi-part TLVs, querying router implementation details, source address verification, signaling aggregate header size limits, advertising multi-interface ID for L2 bundle members, scheduled metric changes for TBR, colored IP prefix, and using FlexAlgo for SR-based ARPs, each with its own set of considerations and concerns.

## Key Discussion Points

*   **Multi-part TLVs:** The draft is proposed for working group last call, emphasizing the need for implementations to provide controls to prevent or deal with partial deployment.
*   **Router Implementation Details:** A draft aiming to query router implementation tires of protocols was discussed. The working group agreed to adopt it.
*   **Source Address Verification:** A method for source address verification in intra-domain networks was presented, which led to a discussion on protocol extensions.
*   **Signaling Aggregate Header Size Limits:** A proposal to signal aggregate header size limits via IGP extensions sparked concerns about the IGP acting as a "dump truck" and the complexity of IPv6 extension headers. Suggestion to move discussion to 6man WG.
*   **Advertising Multi-interface ID for L2 Bundle Members:** A draft that would help network operators to control traffic by advertising the local and remote interface IDs for a true boundary member in ITP. The draft will add the other use case beside discovery by the controller.
*   **Scheduled Metric Changes for TBR:** A presentation discussing HP extensions for TBR encountered resistance, with concerns about diverging from the TBR working group's direction (Yang model distribution) and overloading the IGP.
*   **Colored IP Prefix:** Presenting colored IP prefix as an enhancement to IDB's shortcut. The group discussed if tags could also be used for this purpose.
*   **FlexAlgo for SR-based ARPs:** A proposal using FlexAlgo for SR-based ARPs faced scrutiny regarding potential conflicts with existing l2 bundles and the best path metric types.

## Decisions and Action Items

*   **Adopted:** The working group agreed to adopt the draft on querying router implementation details.
*   **Last Call:** The draft on multi-part TLVs is likely to be put forward for working group last call soon.
*   **Hold/Revise:** The draft on FlexAlgo for SR-based ARPs should be put on hold until there's consensus on key issues, and discussions should continue in both LSR and TEAS working groups. The e-bit discussion will be continued on the mailing list.
*   **Mailing List Discussion:** Further discussion on the distribution of schedule information for TBR and whether it should go into the IGP will be conducted on the TBR mailing list, with LSR CC'd.
*   **Update Draft:** Authors of the 'advertising multi-interface ID' document to add another use case that has been discussed in the group, beside discovery by the controller.
*   **Update Draft:** Authors of the 'IGP flash algo with link loss' to add a description stating that it is a good guideline to have a similar method for the Flex Algorithm.

## Next Steps

*   Move forward with working group last call for the multi-part TLV draft.
*   Initiate adoption call for the draft on router implementation details.
*   Continue discussions and revisions on drafts related to FlexAlgo for SR-based ARPs, TBR extensions, and other topics based on the feedback received.
*   Consider operational guidance document on using Link Loss as a Flex Algorithm constraint