IPv6 Working Group Minutes - RIPE 79 Date: 17 October 2019, 09:00 - 10:30 WG Co-Chairs: Jen Linkova, Benedikt Stockebrand, Raymond Jetten Scribe: Nathalie Trenaman Status: Draft 1.Welcome / Agenda Bashing The presentation is available at: https://ripe79.ripe.net/archives/video/241/ There were no questions or comments. 2. Working Group Chair Re-election There were no questions or comments. Raymond Jetten is re-elected as IPv6 WG co-chair. 3. MSIT Ongoing Journey to IPv6-only and Lessons Learned From It Veronika McKillop The presentation is available at: https://ripe79.ripe.net/archives/video/243/ Mikhael Abrahamson commented that it’s great that Veronika is sharing this, he often refers IPv6 sceptics to her talks. He explained that he was so far unsuccessful in requesting Windows Mobile to implement the CLAT code in their mobile phones and asked Veronika’s opinion on that. Veronica replied that Microsoft wants native IPv6 and adds that, as of two weeks ago, Miscrosoft has an Android device. Geoff made an observation regarding the speaker’s assertion that DNS64 is crucial to the Dual Stack strategy. He asked why it is ok to lie in the DNS since there is more push to put validation closer to the edge. He went on to express concerns about the intermediate strategy around Dual Stack relies so heavily on structural lying in the DNS. Veronika replied that, as a network engineer, she has done whatever it takes to make it work for her users. Also, DNS64 is not crucial for Dual Stack, it is for IPv6-only. She avoids it as much as possible. Geoff also asked Veronika if they are avoiding extension headers. Veronika replied that she never thought about it and she will ask the stack team. Jen Linkova made the remark that people who care about DNSSEC and signing the zones should deploy IPv6. She observed that the DNS64 problems would disappear if the destination has a quad A record. Iljitsch van Beijnum commented on Geoff’s remarks on DNS64, as he is one of the co-authors of the DNS64 RFCs. He said that problems occur in the unlikely event that people do validation on the client, but then the client is unaware of the existence of DNS64. Kostas Zorbadelos (Canal+ Telecom) asked if Microsoft looked at software solutions, like MAP-E or MAP-T. Veronika explained that her network is an enterprise network. If they would go for IPv4-as-a-service they could potentially look into it, but not so far. Kostas Zorbadelos asked if Veronika has any preference. She replied no. 4. Types of IP Addresses, Changes from IPv4 to IPv6 and What All This Means for Network Enno Rey The presentation is available at: https://ripe79.ripe.net/archives/video/246/ Iljitsch van Beijnum commented that he is uncomfortable with Ennos statement that Dual Stack is problematic because it scares people away from deploying IPv6. Iljitsch also mentioned that there are private addresses that need special handling and private addresses that don’t and there is private IP address space that needs special handling and there is private space that doesn’t. Iljitsch added that he thought slide 27 had some errors. Jen Linkova suggested that the discussion be taken offline. Friso Feenstra commented on slide 13, saying he wondered whether this also applies if you don’t have a lot of addresses and your company wants to become a digital company and you don’t have an IPv6 strategy to start buying IPv4. 5. Don't let v4 die! Jens Link The presentation is available at: https://ripe79.ripe.net/archives/video/248/ Jen Linkova suggested that questions for this presentation be taken after her presentation. 6. What Shall This WG do? Jen Linkova The presentation is available at: https://ripe79.ripe.net/archives/video/249/ Anna Willson said that NetNews stopped meeting because the coordination requirements went away. Sometimes it is ok that the coordination work of a working group is no longer as pressing as it was. Now it has become a wider topic, it doesn’t require a working group anymore. Fernando Garcia had three comments. First, RIPE decided to have the IPv6 Working Group in the smallest room, that is significant. Second, IPv6 is a sub-developed problem. If IPv6 is not deployed and we end up with NAT over NAT over NAT, the Internet as we know it will disappear. His final point was that this working group might turn into a self-help group that offers psychological support. Nico Schottelius had two comments. First, with regards to IPv6 deployments there are only two things to fix. The connectivity and the content. If we fix these two, we are done. Some of us are network operators, we can influence our work every day. He said he has a more radical opinion on the content side, according to which we can use our spare time to build IPv6-only services and games. Lee Howard said he liked the suggestions that were made and volunteered to make the business case, as he has been doing this for many years at NOGs. For the hackathons, collecting open source projects and create IPv6-only content, he offered to provide a Google-sheet with a list of projects that need contributions. Jen Linkova said she would be looking forward to the meeting minutes, because there will be some action items. Benedikt Stockebrand commented that we have a killer app for IPv6, called IPv4. 20% of German users are connected over DS-Lite. He added that convincing technical decision makers is the least of our problems, and that work also needs to be done convincing our own direct colleagues. Jen Linkova thanked all participants and closes the meeting.