[ipv6-wg] Request for approval of the minutes for the session at RIPE-60

Please find attached the minutes of our last WG session, which was held during the RIPE-60 meeting in Prague.

The minutes will be declared final if no objections, except for minor editorial corrections that will be fixed before publication, will be received by the end of Wednesday June 16th, 2010 in a timezone of your choice.

Thanks to Vesna and Amanda from the NCC for scribing and preparing these minutes. Feedback is welcome on this list or if you prefer a less public method via ipv6-wg-chair at ripe.net.

Greetings,

Marco Hogewoning
-- 

IPv6 Working Group Minutes from RIPE 60
Status: Draft
Revision: 1.1

WG Co-chairs: David Kessens, Marco Hogewoning

Scribe: Vesna Manojlovic (RIPE NCC)
Date: Tuesday, 4 May 2010,
Time: 16:00 - 18:00


A. Administrative items

Introduction
IPv6 co-Chair, David Kessens, welcomed the participants.

New Chairs
David introduced two new co-chairs: Marco Hogewoning (XS4ALL) & Shane Kerr (ISC). The new email address to reach all three co-chairs is: Ipv6-wg-chair at ripe.net

New Charter
David discussed that following RIPE 59 in Lisbon, a new charter was suggested and discussed on the mailing list. A consensus was reached on the mailing list. The charter has four focal points: outreach, education, sharing deployment experiences, and discussing and fixing operational issues.

RIPE 59 Minutes
Minutes from the previous WG meeting were published to the list.  There were no comments and the minutes from RIPE 59 were approved.

B. Java script-based HTTP Client Measurements  (IPv6 at Web Clients and Caching Resolvers)

Emile Aben from the RIPE NCC gave a presentation on JavaScript-based HTTP Client Measurements. The presentation is available at: <http://www.ripe.net/ripe/meetings/ripe-60/presentations/Aben-IPv6_at_web_clients_and_caching_resolvers.pdf>

Link to the articles on the RIPE Labs:
<http://labs.ripe.net/content/measuring-ipv6-web-clients-and-caching-resolvers-part-1> 
<http://labs.ripe.net/content/measuring-ipv6-web-clients-and-caching-resolvers-part-2-country-level-and-other-statistics>
<http://labs.ripe.net/content/measuring-ipv6-web-clients-and-caching-resolvers-part-3-methodology>

C. IETF v6ops activity

Kurt Lindquist from Netnod gave a presentation on the IETF v6ops activity. The presentation is available at:
<http://www.ripe.net/ripe/meetings/ripe-60/presentations/Lindqvist-IETF_v6ops_Activity.pdf>

IETF v6ops WG does not produce standards; produce documentation on BCP. We invite everyone to join the mailing lists discussions.

D. CPE Overview

Marco Hogewoning from XS4ALL gave a CPE overview. The presentation is available at:
<http://www.ripe.net/ripe/meetings/ripe-60/presentations/Hogewoning-IPv6_CPE_overview.pdf>

Remco van Mook said Linksys does 6to4. It is undocumented. Lookup the presentation from NANOG from last year.

Marco asked for a request for feedback, if there was any interest and if anyone would volunteer for testing. 

Lars mentioned that there was one Marco didn't mention: Mikrotik 5.0 beta is out, supports PPPoE.

Marco asked if it didn't do ADSL interfaces. 

Lars said that was indeed the case. 

Wolfgang Tremmel from DE-CIX asked who is using which CPE. 
AVM 4
Cisco (a lot)
Lynsix ?
Native?
Tunnels?

Daniel Karrenberg  from the RIPE NCC stated they were interested in publishing this, not only from meeting to meeting, but continuously.  He then suggested doing a similar survey for the pre-packaged firewall deal.

E. IPv6 Worldwide Policy Review
Filiz Yilmaz from the RIPE NCC gave a presentation on IPv6 world-wide policy review. The presentation is available at:
<http://www.ripe.net/ripe/meetings/ripe-60/presentations/Yilmaz-IPv6_WorldWide_Policy_Review.pdf>

Filiz said that policy is not normally discussed in the working group, but that they want to make sure people are up-to-date with the policy. She said that people are deploying v6 now, and the policy was written a decade ago.  She added that policy is trying to catch up with the operational experience and that there were two main issues: site definition, and routing practices. Filiz mentioned that there is a new article published on the differences between v6 and v4 PI policies.

F. Mobile broadband and IPv6 in Slovenia
Jan Zorz from go6.si gave a presentation on IPv6 in Slovenia. The presentation is available at:
<http://www.ripe.net/ripe/meetings/ripe-60/presentations/Zorz-Mobile_broadband_and_IPv6_in_Slovenia.pdf> 

Marco Hogewoning asked if it was capable of both IPv4 and IPv6 or if you had to choose.

Jan replied that it could do both, but it makes one and then closes down the PPP and makes another one. He said that at the same time, on the same phone, you couldn’t have a look at two different websites on both v4 and v6.

An attendee asked if there was roaming.

Jan replied that Tusmobil was very small and didn’t cover the whole country, but that Mobitel had roaming.

Marco Hogewoning mentioned a last-minute agenda item: Martin Levy from Hurricane Electric would present for a few minutes on their BGP Toolkit.

X. Hurricane Electric’s BGP Toolkit

Martin Levy from Hurricane Electric gave an update on BGP Toolkit. There was no presentation, but more information can be found at:  
<http://Bgp.he.net>

Martin said they do daily processing of RIS & Routeviews data. You can do searches on ASs, prefixes, etc.  He added that everything was v4 & v6 related.  He said they  added the “graphing”. (as1273).  He mentioned that it gives you the ability to see how are customer routes seen from the rest of the Internet.

Martin asked for feedback.

G. Update on Recent RIPE NCC IPv6 Activity
Chris Buckridge from the RIPE NCC gave an update on RIPE NCC IPv6 activity. The presentation is available at:
<http://www.ripe.net/ripe/meetings/ripe-60/presentations/Buckridge-Update_on_RIPE_NCC_IPv6_Activities.pdf>

Paul Rendek from the RIPE NCC commented that they had a 5 days-long hands-on IPv6 workshop in Bahrain. He said that the next v6 workshops would be in UAE, Jordan, Syria & Istanbul with MENOG.

H. IPv6-wg brainstorming
Marco Hogewoning then opened up a brainstorming session on IPv6. He mentioned several issues to consider: Where do we go from here? Where should we focus on?
Outreach? New audiences or areas? Education? Do we need any? What do we need?Cooperation ? Platform for technical issues? Other organisations?

Kurt Lindquist from Netnod mentioned that at a Netnod customer meeting, Ole told us about Broadband forum and standardisation. He said that more of that would be outreach and education at the same time.

Marco asked that if anyone knew of people that deployed IPv6 to come to the IPv6 Working Group and share their achievements.

Jan Zorz commented that people needed to start ‘walking the walk’. Find people who can present stuff that is actually working. He commented on the need for more operational talks.  

Geoff Huston said that very few folks who run big networks want to do any research about v6. He asked how can we take the measurements tools and put them on the high-traffic sites? How do we engage in the useful measurements? If the 5% is the real number of End Sites that are IPv6-capable, then we have reached the goal. But is the number really 5%?

Marco commented that the measurements could be done on the tax-office sites. That gives you Joe Average. 

Geoff replied that 1.6% path MTU failure rate wasn’t good enough, that it had to be 0%. He said there was a way to do it go for a smaller MTU on the server.

He then said that his other wish was for someone to make Java script that worked in every browser (most importantly in the Explorer), and loads in the background.  He asked for volunteers.

Daniel Karrenberg from the RIPE NCC said that according to the EU Council presentation that morning: “Naming and Shaming works on the country level”. He encouraged people to participate in those measurements, even if the tools weren’t perfect.

He added that he would really like if Emile Aben from the RIPE NCC had 5 to 10 commitments from around our region for the web sites to host our script. He said the RIPE NCC needed to be able to present the data to people who are making business decisions, and national policy makers. 

Kurt Keiser said that IPv6 is about global connectivity and continued connectivity. He asked what happens with v4 clients that can not be upgraded?  He would like to see a listing of devices that can not be upgraded (for example, nuclear power plants).

Kurt Lindquist said it would be good to have a standardised way of collecting data.

He added that there were big web-sites in every county, for example, in Sweden there is Pirate Bay. He said that those guys would be happy to share their statistics. He said he was there was someone in each country to talk to.

Marco said that not everyone wanted to share how many visitors they had. But  he hopes they can show a percentage of usage [over IPv6] without showing numbers.

Jamie Stallwood said that the IPv6 Ripeness presented today was interesting. He added that it would be good if the info could distinguish between edge ASs and the big, transit ones. Also the regional overview would be interesting. For example, in UK the big transit providers don’t do v6.

Emile Aben from the RIPE NCC said the suggestions would be considered.  

Alex Band from the  RIPE NCC said that he’d talked with hundreds of people as a Trainer and that  there didn’t seem to be one place with practical information for small ISPs. What are the steps? Practical experiences?  He said that once they had experience, they would share it.  Alex said a platform needed to be created to do that.

David added that such information was available in 6bone Database, like where to ping to check if the things are working. He said that was all gone now because the 6bone didn’t exist any more.  

Alex Band from the RIPE NCC mentioned a presentation Marco Hogewoning did about XS4All’s practical information about their network setup, some tricks and tips, and that when it was showed during RIPE NCC training courses, it was really appreciated.

David asked if he was looking for the internal information.

Marco said he received good feedback about his slides, and Mirjam Kuehne from the RIPE NCC said that they might add an IPv6 section on RIPE Labs.

Marco made a request to the community:  If you have those deployment scenarios and are willing to share your address plan,  share the data by sending the link to the mailing list.

Aaron Hughes from 6connect, Inc., said they are working with ARIN and NANOG on a site where authors and readers can manage living documents, with an interface that allows you to search through documents. The mailing list is: Bcp-discuss at bind.com; there will be a BOF at NANOG 49 to agree on this.

Marco mentioned that he is also working on a registration need for IPv6 space.  He said it would be discussed in the Address Policy WG.

David Kessens asked the audience if it wanted to set up the IPv6 meetings in a different format, like tutorials or a Plenary or a v6-only day. He asked the attendees to give feedback.

Marco added that there would be one more global IPv6 survey.  He instructed the audience to approach Maarten Botterman from GNKS Consult if you anyone was interested. He said the survey would be sent to the IPv6 WG mailing list.

Y. Input for the RIPE NCC Activity Plan

Z. AOB & wrap-up

Marco ended the session.