RIPE 90 - RIPE Community Plenary Minutes
Thursday, 15 May 2025
16:00 – 17:30 (UTC+1)
Chair: Mirjam Kühne
Scribe: Ulka Athale
Status: Draft
1. RIPE Chair Team Report
Mirjam Kühne, RIPE Chair
RIPE Chair Mirjam Kühne welcomed attendees to the session and shared the work of the RIPE Chair team since RIPE 89. The RIPE Chair team had declared consensus on ripe-838, a document that describes the relationship between RIPE and the RIPE NCC. They also worked on reviewing the entire policy portfolio with the RIPE NCC Policy Officer Angela Dall’Ara and on the consistency of metadata in the RIPE Document store. Additionally, the Chair team has been continuing their outreach efforts and monthly update on RIPE Labs.
There were no questions or comments.
2. NRO NC: Report from the ICP-2 BoF
Hervé Clément, Andrei Robachevsky and Constanze Bürger, NRO NC
The NRO NC representatives shared an update on their efforts to gain input on the draft “RIR Governance Document” intended to update ICP-2. They shared key takeaways from the BoF session on Monday, hallway conversations and their presentation during the Address Policy Working Group session. RIPE community feedback focused on the issues on unanimity in decision-making, the assumption of good faith by all parties, whether there needs to be a core set of features common to all registries, the importance of business continuity and some bigger picture comments on the role of RIRs in the Internet ecosystem.
Remco van Mook thanked the NRO NC for their work. He pointed out that this work takes place in the context of an existing system. He expressed concerns with the writing of the document, particularly fiduciary duties and consensus between RIRs. It is very important that the Boards of the respective RIRs will be able to sign off on this document in line with their fiduciary duties once it is complete.
Lu Heng, Larus Limited, made a comment on number portability and the recognition of member rights. He stated that his law firm has begun a potential anti-trust investigation that the RIPE NCC could get out of simply by recognising IP ownership rights officially policy documents, corporate documents and in ICP-2. The current model is dangerous and unprepared for absolute ownership as the databases have over 200 billion dollars in market value. He urged the ASO AC to take a serious view and look at the governance changes for the 25 years ahead. He said that he would write to the RIPE list about the anti-trust matters.
Tobias Fiebig reminded the room that this discussion about the value in the database stems from an artificial shortage of IPv4 resources which could be alleviated by migrating to IPv6. He stated that anyone acting in good faith would ask for an IPv4 flag day and not for any other approach to solve this problem.
3. Update from the RIPE NomCom 2025
Jan Žorž, RIPE NomCom
Jan explained the process of selecting the NomCom and the non-voting liaisons supporting them. He introduced the candidates for RIPE Chair and Vice-Chair and how the NomCom had kept office hours during RIPE 90 to hear community feedback.
Brian Nisbet, HEAnet, said that half blaming the community for the lack of diversity of the NomCom is incorrect. We need to review the process and work as a community to make sure that people are happy to volunteer and commit to it. He said that it was possible to design a process to support a diverse NomCom and volunteered to support this process.
Jan said that there will be a NomCom report with advice for the next NomCom which could create a community discussion as there needs to be a change in both the policy and accompanying process. He thanked Brian for his comment.
Leo Vegoda, And Polus LLC, said that someone had asked him earlier about voting in the NomCom, which highlighted a gap in communication. This could also have been a factor feeding in to the lack of diversity of people putting themselves forward for the NomCom.
Jan pointed out that there was a presentation in the Opening Plenary at RIPE 89 in Prague and there were also extensive explanations.
Daniel Karrenberg said that there was some truth to Leo’s comments, there had been extensive communication over the last few meetings about how the NomCom works and that he had personally canvassed for people to volunteer. The willingness of people to come forward had been less than stellar.
Niall O’Reilly, RIPE Vice-Chair, stated that it would probably be an early item on the agenda of the new RIPE Chair team and clarified that he was not running for the position again.
Jan thanked everyone for their suggestions and added that the RIPE NomCom process was a lot of work, not just for himself as the Chair but for everyone involved. He thanked the NomCom for their work.
4. The Next Generation of BGP Data Collection Platforms
Thomas Holterbach, University of Strasbourg
Thomas Holterback presented on the next generation of BGP data collection platforms that combines route collectors to provide a larger view of Internet routing. He invited the community to try it out and share feedback.
Sander Post, Ventus asked about the collection and high coverage. He asked if they would end up with the same features as other tools.
Thomas replied that it was possible, 90% of the time two routers in the same ASN would see the same things but they cared about the one person that is different. This is why he encouraged others to peer with them even if they share data with other tools.
Antonio Prado, Fiber Telecom, asked about the long-term business model behind this project.
Thomas replied that for now it’s a research project but they hope to make it sustainable in the long run so they would need to find funding. They are also working on reducing costs.
Tobias Fiebig, MPI-INF, asked about the difficulty in obtaining example data from routeviews since the database is so vast. He asked about whether they had considered ingesting all that data into their system given that it’s more space efficient.
Thomas said that it was possible but there is also a lot of data in real time, roughly 1 gigabit a day. At some point it will be tough system-wise to populate the data in real time in addition to the past data from route views.
Tobias commented that if they were currently facing performance issues they would need to ingest route view data at roughly 20 times the speed of real time data. This means they would need a year or two to ingest the whole routers.
Thomas replied that they had an algorithm optimised for this but they were also running on two community servers that have a 2 terabyte disk capacity with limited funding. At present, they have too many hardware capacity and funding constraints.
Tobias asked if it would be possible to have a large influx of peers at present.
Thomas clarified that real time peers were not a problem, but the issue lay with historical data. They would not have the capacity to take all the real time and historical data with their existing hardware.
Emile Aben, RIPE NCC, said that he was involved in making the peering policy a bit more selective because they got very similar peers, so volume was not the main concern, but similarity. He asked whether this simulation could help the RIPE NCC find peers that add diversity to the platform?
Thomas asked about the definition of similarity and pointed out that someone might be interested in measuring similarity between vantage points. He said that their approach was different because there could be things of interest to someone, for instance, whether a prefix is visible from everywhere. In their SIGCOMM paper, the approach they chose did not involve testing various use cases to assess whether a new vantage point would be more effective. Instead, they adopted a method similar to image compression. Just as image compression retains only the pixels necessary to reconstruct the original image, their goal was to discard data from the BGP archive that was not essential for reconstructing the archive. He also noted that the algorithms they developed were published in the paper, made open source, and are available online. This allows others to use the tools and determine which parts of the data are significant, which could be useful for anyone trying to decide whether the information is relevant for their needs.
5. The new RIPE Meeting website
Phillip Oldham, RIPE NCC
Phillip introduced the new RIPE Meeting website and explained how the different systems are being consolidated into a single tech stack. The current meeting presentation submission system is being replaced by an open source tool called Pretalx, which is used by many NOGs and other technical communities. He invited the community to share feedback with the RIPE NCC Web Services team.
There were no comments or questions.
6. AOB and Open Mic
Daniel Karrenberg clarified that he was not against reviewing the NomCom process. His remark was directed towards those who felt that not enough effort had been made to make the NomCom more diverse and inclusive.
Tobias Fiebig commented that the most important takeaway was that there was an issue and the community needed to revise their approach to the NomCom process and strive to do better.
End of session.