SIDROPS at IETF-121
Session 1: Tuesday, November 5, 2024 16:30-17:30 Local Time in Ireland
(GMT+1)
Room: Wicklow Hall 1
Session 1: Tuesday, November 5, 2024 16:30-17:30 Local Time in Ireland
(GMT+1)
Room: Wicklow Hall 1
2) Tom Harrison - [10 min]
draft-ietf-sidrops-manifest-numbers-02
Tim - Useful to have recommendations on publications
Tom - Yes
Russ - Should be done before WGLC?
Tim - Sure
3) Sofía - [15 min]
NRO RPKI Program Update
4) Job Snijders - [15 min]
Requirements for a next-gen RPKI transport
Rob - RRDP is transactional compared to RSYNC. RRDP has implmnetation
errors than protocol. Supportive of Transport Requirements
Job - Thanks
Rudiger - RSYNC has vulnerability
Job -
Benno - 2 suggestions - Supportive of Transport requirements. Keep WG
document.
Job - No goal of RFC yet, but publish for future developers
Agree with complexities.
Andrew - Requirements document is a good thing. Have u considered fixing
issues of RRDP?
Job - Appreciate the feedback. Deflates used by many parties for
measuring convergence. RRDP desync is present. Some stuff is brought
into RRDP but
Maria - Fix issues by putting it in GIT
Warren - Publishing into RFC is better.
George - Requirments is a good thing to do.
Jeff - Putting in BGP is not a good idea. Out of band is a better idea.
Don't refer IDR for reference for requirements.
Job - Thanks for the feedback
Keyur - Agree with Jeff. Do we have doc for different implmentations of
RRDP.
Job - NO implmentations info available
5) Job Snijders - [15 min]
draft-snijders-constraining-rpki-trust-anchors-06
Warren - AREN doesn't have IPv6 transfer Policy
Job - People might configure something and doesn't touch it
Rudiger - Resource record belongs and what not for someone will help in
checking consistancy.
Ben - Informational RFC is a good idea.
Tim - Aware of this.
Session 2: Tuesday, November 5, 2024 18:00-19:00 Local Time in Ireland
(GMT+1)
Room: Wicklow Hall 1
1) K. Sriram - [10 min]
draft-geng-sidrops-asra-profile-00
2) K. Sriram - [10 min]
draft-sriram-sidrops-asra-verification-00
Ben - All analysis is correct. Don't looks like a problem to be solved
since providers don't lie to each other.
BGPsec solves fake link problems
SriRam - Partial deployment, there is a customer attacking the provider.
Ben - BGPsec ccan still ahndle it
SriRam - We can dicuss
Alexander -
SriRam - Transpernt RS: Client can list all of its client
Alexander - Universal API
SriRam - RS can list the clients. It is starting point and can be
refined
Warren - ASPA should be published first
SriRam - Sure
Antoin -
3) Libin - [5 min]
draft-chen-sidrops-sispi-02
4) Shuhe Wang - [10 min]
draft-wang-sidrops-route-partial-visibility
5) Shenglin - [10 min]
draft-jiang-sidrops-psvro-00
Job - What is limited coverage in RoA? all prefixes or Multiple ASes?
Please clarify
Shenglin - Will respond to it later
6) YingYing - [10 min]
draft-li-sidrops-rpki-repository-problem-statement-00
Mingqing - Agree
Chris - Does the work listed here mentions that redundant RPs doesn't
work? No need of more specification rathen than how to operate.
YingYing - Lets take it to mailer
Alexander - which reasons where these polls were made?
YingYing - globally. Randomly chosen.
7) Jia Zhang - [5 min]
draft-zhang-sidrops-aspa-egress-00