IPv6 Working Group Minutes - RIPE 80

Date: 14 May 2020, 13:00 - 13:45
WG Chairs: Benedikt Stockebrand, Jen Linkova, Raymond Jetten
Scribe: Olivia Ruimwijk
Status: Draft

1. Welcome - Agenda Bashing

Presentation available at:
https://ripe80.ripe.net/archives/video/362/

There were no questions or comments.

2. Why Is It so Hard to Implement IPv6?

Wilhelm Boeddinghaus

Presentation available at:
https://ripe80.ripe.net/archives/video/363

Chris Conway (Conway TV) asked the presenter about his experience using IPv6 with Microtek.

Wilhelm Boeddinghaus replied that he had no experience using Microtek.

Aleksi Suhonen (Trex Regional Exchanges) asked why the lock files should be doubled.

William clarified that the file entries and not the file itself should be double locked. He added that with firewall rules for HTTPS, IPv4 and IPv6, you get lock file rules with both protocols and it makes it harder to read, evaluate and find the attacks.

Jordi Palet Martinez (Moremar, S.L.) commented that Microtek was the worst IPv6 vendor as they only used dual stack and had the wrong naming for protocols. He added that he reached out to Microtek to offer his help but that they never replied to him.

Wilhelm pointed out that the technology was getting better even if a vendor like Microtek still has problems to solve. However, he added that IPv6 inertia was still commonplace within organisations and that better documentation could help addressing this issue by making things easier for administrators.

Maximilian Wilhelm (SDN.clinic) asks the presenter if he had a solution regarding servers which need access to IPv4 services as it won't work with dual-stack and IPv6-only it's still problematic.

Wilhelm replied that you can implement dual stack when you need both, IPv4 and IPv6, in parallel. However, he added that as long as you insist on dual stack everywhere, you don't get any benefits from a new network design.

Christian Bretterhofer (Andritz AG) asked to compile a list of compatible IPv6-only equipment.

Wilhelm answered that it was possible and asked Christian to follow-up on the WG mailing list.

Falk von Bornstaedt (RIPE NCC Executive Board) asked if it would be possible to set IPv4 end dates for certain regions, such as mass volume streaming and if we could work on a common phase out plan.

Wilhelm replied that there was a similar effort from the sunset IPv6 Working Group at the IETF, but he believed that this group was not active anymore. He pointed out that setting up IPv4 end dates might clash with commercial interests as companies like Facebook or Google could lose a big part of their users in the process.

Wilhem connection was interrupted and the Chairs asked attendees to send remaining questions to the mailing list.

3. IPv6 SLAAC & Renumbering Events

Fernando Gont

Presentation available at:
https://ripe80.ripe.net/archives/video/365

A participant commented that reducing the valued LFD may end up in host W/O with any IPv6 prefix, so that even local communication is no longer possible. He added that if the router disappears, you will need another one to switch to, otherwise you are offline until then the router comes back on. He concluded that we need routers which are able to withdraw prefixes that are no longer used.

Fernando answered that in a home deployment scenario, the router is the same device that you use as a switch. This means that if the device is not working, you won't be able to do much. He recommended to get unique local addresses (ULA) to get addresses that are stables and don’t have prefixes that depend on your connection with the ISP.