RIPE 56

• content to enum-wg-chair@ripe.net
• format to webmaster@ripe.net

RIPE 56 ENUM WG Minutes
Thursday 8 May, 2008, Berlin
Chair: Niall O'Reilly
Scribe: Franz Schwarzinger
Jabber: Chris Buckridge

Webcast and Feedback Archives:
http://www.ripe.net/ripe/meetings/ripe-56/presentations/thursday.html

A. Administrative Matters

• Welcome
• Select a scribe
• Jabber Monitor
• Microphone Etiquette
• Finalise agenda

B. Minutes from RIPE 55

Niall O'Reilly advised that the minutes had already been approved on the WG mailing list.

C. Review Action List

54.3 [Niall O'Reilly]: Communicate result to WG Chairs, RIPE-NCC

55.1 [Niall O'Reilly, Carsten Schiefner]: Update the RIPE 54 minutes to take out the typos.

55.2 [Niall O'Reilly, Carsten Schiefner]: Take account of the offer of the Crawler presentation in planning WG session for RIPE 56

All actions have been closed since the ENUM WG at RIPE 55.

D. Main Presentations

D1: Crawler Tool

Alexander Mayrhofer (enum.at)
http://www.ripe.net/ripe/meetings/ripe-56/presentations/Mayrhofer-The_ENUM_crawler.pdf

Comments:
Ed Lewis (NeuStar) suggests that some people might use wild-cards for over-dialling.

Alexander notes that one ISP in Austria is using wild-cards for normal numbers, even without having extensions behind that. It's possible for them to change the statistics significantly.

Carsten Schiefner (Deutsche Telekom) asks about the reason for the high number in the UK.

Alexander thinks that the reason for this is the carrier registration of ENUM in the UK. He asks the community for feedback and suggestions regarding his system.

D2: +31 in Production

Antoin Verschuren (SIDN)
http://www.ripe.net/ripe/meetings/ripe-56/presentations/Verschuren-ENUM_in_NL.pdf

There were no questions.

D3: +44 Coming soon

Denesh Bhabuta (UK ENUM Consortium)
http://www.ripe.net/ripe/meetings/ripe-56/presentations/Bhabuta-UK_ENUM_Happenings.pdf

There were no questions.

E. ENUM Operations

E1: Tier-0 Report

Anand Buddhdev (RIPE NCC)
http://www.ripe.net/ripe/meetings/ripe-56/presentations/Buddhdev-RIPE_NCC_ENUM_Update_Tier-0_report.pdf

Peter Koch (DENIC) thanked Anand for the update and asked whether the RIPE NCC has monitored the amount of DNS queries that are actually soliciting DNSSEC responses and whether this has changed recently or not.

Anand responded that they have not started monitoring this yet, since they just got their first secure delegation last week. He will provide more information on that at the next RIPE Meeting.

Ondřej Surý commented that he did a secure delegation as well and actually did this one day before the Polish and were therefore the first ones.

Niall O'Reilly (Chair) comments that point X on the agenda is now also covered. This point would have been the broadening of the DNSMON criteria through the RIPE NCC.

E2: Tier-1 Performance/Availability

Carsten Schiefner (Deutsche Telekom)
http://www.ripe.net/ripe/meetings/ripe-56/presentations/Schiefner-Tier-1_Performance_Availability.pdf

Olaf Kolkman (NLnet Labs) asks if this is an action on Carsten personally or as a Co-Chair?

Carsten said that he wants to take this as a personal action, as an interested member of the ENUM community.

Jim Reid commented on the political implications of this proposal, stating that it would be most inappropriate for the RIPE NCC to ask hosts of lame name servers to fix it or even pull their delegation. It should only be presented as a statement of fact to the ITU since it is a purely operational thing.

Carsten asked if he would suggest a best practice document or an informational recommendation.

Jim does not think it needs to be that elaborate. A letter, written in such a way that it does not start sending off alarm bells, informing the delegates and other members should be sent.

Niall (as chair) proposes to put an action on this.

Jim said that the RIPE NCC should do this.

Olaf stated that the RIPE NCC is not in the position to contact the ITU about this. However, such a document should pass the RIPE NCC and the IAB before it is passed on to the ITU-T.

Niall comments that Olaf should see this text as well.

An attendee expressed his interest. He asked which path should be followed in order to solve this problem. He compared it to the Youtube hijacking. People would start hard-coding these countries in their software that should not be queried due to lame delegations. It might lead to problems in the future.

Robert Schischka (Austrian ENUM registration) said that Non-working delegations are not very good for ENUM eco-system. Lame delegation and every failure or stability problem will reduce the general trust in the system. He also considered a distributed block list, which is already used by some Austrian ISPs, a bad thing. Every holder of a delegation should see
how it sees itself and maybe consider its service best effort instead of production ready.

Carsten said that the question whether a service should be in production or not might be covered by the next presentation.

Patrik Fältström (Cisco) stated that they need to be very careful. If somebody does not manage their delegation, we will not be able to know if a call actually reaches the target. This group should be very careful about what they decide on this issue. It might be difficult to get people's trust with lame delegations in place.

Jim agreed that they need to be very careful about the wording in this document.

Peter agreed with Patrik, and said that screwing up one country, does not mean they screw up an other country as well, but it could potentially lead towards ENUM losing trust. He didn't see this as a big road block. It's explicitly not about chasing down every lame delegation. He wondered if these cases even happen enough to make coming up with a policy document or a heavyweight proposal even worthwhile. Maybe one can deal with this quickly on a case by case basis.

Carsten Said that it is always a matter of definition and this happens on a regular basis.

Peter asked if these are repeat offenders?

Carsten responded that there were repeat offenders.

Peter stated that this could be dealt with.

Carsten answered that this is the reason why this is being discussed here and they need to come up with a solution.

Peter proposed to deal with repeat offenders on a case by case basis.

Niall proposed to put an action item on Jim to produce a draft guidance document for the RIPE NCC regarding operational problems. The document should be lightweight but supportive for the RIPE NCC. It should be written by this working group with Jim as editor.

Niall noted a new action point 56.1 on Jim Reid.

Patrik commented that having this document would be a good idea, but wondered, due to the global role of the RIPE NCC in the ENUM sector, if this document should be written by a higher instance. He advised the working group to be careful.

Niall agreed with this. Due to the heavy implications on layer nine and above the document should be top-heavy but still lightweight.

Olaf commented that if the document would actually be necessary if it is already supposed to be so lightweight.

Niall stated that repeat offending has been difficult to deal with in the past. Case by case hasn't quite done it, so they need to
document this.

Peter said he would like to see these operational problems a bit more qualified. He thought that the RIPE NCC can be trusted to deal with this. He also stated that they are actually not after these single lame delegations.

Carsten added that Peter should be part of the team that will write this document.

Patrik noted that the working group should think about who this document will be addressed to. Is it only the RIPE NCC or IAB as well?

Niall agreed and suggested to move this discussion to the mailing list.

F. Short News

F1. enumdata.org Update

Carsten Schiefner (Deutsche Telekom)
http://www.ripe.net/ripe/meetings/ripe-56/presentations/Schiefner-enumdata.org_update.pdf

Y. A.O.B.

There was no other business.

Z. Close

There were no action points open at the beginning of the meeting.

One new action point was opened at the meeting, as follows.

ENUM-AP-56.1: [Jim Reid]
Edit guidance document regarding operational problems.
Carsten Schiefner, Niall O'Reilly, Patrik Fältström, Peter Koch, and Olaf Kolkman will also be involved.

Everyone was thanked for their participation.

End.