**Session Date/Time:** 25 Jul 2023 00:30 # sidrops ## Summary The SIDROPS meeting at IETF 117 in San Francisco covered several topics, including requiring implementations before draft advancement, a new draft on prefix lists, updates on ASPA drafts, and a proposal to improve the efficiency of failing over from RRDP to rsync. There was also discussion regarding a new approach for total address with addition security and privacy ## Key Discussion Points * **Implementation Requirements for Drafts:** Job raised the question of requiring implementations before drafts with software implications are advanced. Jeff expressed concerns about the resources required and the potential for hindering early innovation. Russ agreed to write a mail to discuss further. * **Prefix List Draft:** Jeff presented a draft regarding prefix lists and their use with ROAs, acting as a signed route object. Warren questioned the utility of the draft, suggesting it might not solve the intended problem of AS impersonation. Rudiger clarified that it protects against a different AS announcing the stuff. Ben and Joe express support for the effort. * **ASPA Draft Updates:** Sridhar provided updates on the ASPA verification and profile drafts, noting the removal of the AFE limit and subsequent adjustments. He raised questions about egress eBGP ASPA verification and its complexity. There was discussion about potentially focusing on using the OTC attribute (RFC 9234) instead. * **Validated RPKI Data Notation Draft:** Oliver presented a draft proposing a BNF notation for displaying validated RPKI data, intended for informational purposes to simplify testing and documentation. Jeff suggested considering Yang module integration and sorting. * **Publication Server Best Practices Draft:** Main and code presented a draft outlining best practices for running a publication server and rsync repositories, aiming to document lessons learned and improve repository reliability. * **RRDP to Rsync Failover Efficiency:** Joe presented a mechanism to improve the efficiency of failing over from RRDP to rsync by aligning file system timestamps with internal RPKI timestamps. Discussion focused on the choice of timestamp (CMS signing time vs. notBefore) and the importance of a consistent approach. The group leaned toward CMS signing time to align with the flow of data and encourage a purpose for the attribute. * **Aspire Objects Profile Specification Update:** The working group fixed the the ASR 1 text encoding. There was discussion about whether transit freeness should be expressed by using an integer of value 0 in in the s ASID. or use an empty sequence where providers simply are absence. ## Decisions and Action Items * **Implementation Requirements for Drafts:** Russ will start a mailing list thread to discuss the proposal to require implementations before draft advancement. * **Prefix List Draft:** Jeff will review the DEEPE draft by Rudiger and investigate similarities * **ASPA Drafts:** Authors (Sridhar, Ku, Job) will discuss the inclusion of egress eBGP ASPA verification and RFC 9234. * **Validated RPKI Data Notation Draft:** Oliver will consider sorting in the syntax and CSV format. * **RRDP to Rsync Failover Efficiency:** The working group will work to come to consensus on which timestamp to use from the RPKI object and ensure that this is documented. * **Address with addition proposal:** the general sense was that the team should progress the work in a group more specialized on security, with a potential review by sidrops when more mature. ## Next Steps * Continue discussion on the mailing list regarding implementation requirements for drafts. * Authors will refine the ASPA drafts based on the discussions. * Working group adoption call will be made for the validated RPKI notation drafts * Authors will progress the the "Address with addition" proposal in a better suited WG (likely ipsecme) and request a sidrops review when more mature.