Markdown Version | Session Recording

Session Date/Time: 05 Nov 2024 15:00

lisp

Summary

This LISP working group session covered several key documents and proposals, including updates on the LISP DTT document, a new proposal for LISP in Savvy context, and progress on multicast documents and the YANG model. Discussions focused on technical details, inconsistencies, and potential areas for improvement. The group discussed action items related to document updates, security reviews, and further analysis of the interplay between different LISP features.

Key Discussion Points

• LISP DTT Document (RFC 8111 bis):
  • Inconsistency in the definition of the instance ID (IID) - whether it should be 24 or 32 bits. The current on-the-wire representation is 24 bits.
  • Need for revision of pseudocode and examples in the document.
  • Definition of DTT referral cache is still pending.
  • Signature algorithms used in the document should be reviewed for compliance with current security recommendations. Specifically, alternatives to RSA (e.g., elliptic curves) should be considered.
• LISP in Savvy Context:
  • The draft proposes using Savvy procedures to provide source address protection in LISP networks.
  • Discussion on whether the proposed mechanism duplicates existing functionality in the EID mobility draft. Concerns raised about potential mobility latency introduced by Savvy.
  • Question regarding how the mechanism handles IP address reclamation when a legitimate user goes away and a spoofer takes over.
  • Discussion on how the proposal interacts with anonymacy EIDs (CID).
  • Dino inquired about having one EID associated with different MAC addresses roaming. Savvy does not support this scenario.
• LISP Multicast Documents (RFC 6831 bis, RFC 8378 bis, Group Mapping Draft):
  • Request to move all three documents to Last Call.
  • Discussion on the destiny of the group mapping draft - recommendation for experimental status.
  • Clarification on the difference between RFC 6831 and RFC 8378 (signaling vs. non-signaling approaches).
  • VDAS document with hashing is suitable to be included in the standard track.
• LISP YANG Model:
  • Updates to the YANG model based on previous review feedback.
  • Discussion on how to handle existing use case informational documents, mainly for the case of the satellite use case to be in sync with other use cases like the NAC one.

Decisions and Action Items

• LISP DTT Document:
  • Address security review of the signature algorithms. Ad hoc security review by security directorate.
  • Decide the bitlength of the Instance ID, probably it's going to be 24 bits.
• LISP in Savvy Context:
  • Mark will send the document on the mailing list.
  • Mark and Dino will meet to discuss the overlap with the EID mobility draft.
  • Explore to consider this case for the first confessor in Savvy and document the way to solve this.
• LISP Multicast Documents:
  • Chairs, along with volunteers, will review all three documents before initiating Last Call.
  • Consider making the hashing component of the VDAS document part of the standard track.
  • Dino will reach out to the NAC author to proceed to continue the process.
• LISP YANG Model:
  • Submit a new version of the YANG model incorporating the discussed changes.
  • Submit directly to the working group Last Call once addressed the minor things left to fix.

Next Steps

• Mailing list discussions and offline meetings on identified open issues.
• Document revisions and resubmissions.
• Security directorate review of the LISP DTT document.
• Working Group Last Call for the LISP multicast documents.
• Continued progress on the LISP NAT traversal and decent drafts.