RIPE 58

RIPE Meeting: 58
Working Group: ENUM
Status: Final
Revision Number: 1

 

RIPE 58 ENUM WG Minutes
Thursday, 7 May 2009
Chair: Niall O'Reilly
Scribe: Nathalie Trenaman

Webcast and Feedback Archives:
http://www.ripe.net/ripe/meetings/ripe-58/archives.php?day=thursday

A. Administrative Matters

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

B. Minutes of ENUM Working Group at RIPE 57, Dubai

The minutes were approved.

C. Review Action List

56.1 [Jim Reid] Edit guidance document regarding operational problems

Jim said that he had still not received any input about this item. The action item was declared closed.

D. Main Presentations

D1: RFC 5483: ENUM Implementation Issues and Experiences
Lawrence Conroy
http://www.ripe.net/ripe/meetings/ripe-58/content/presentations/enum-experiences.pdf

Patrik Fältström said that he thought that Lawrence was mixing up problems with quoting in provisioning systems with quoting in the DNS. He said that he thought that it was important that these things were kept separate.

Lawrence responded that he agreed and did understand the difference. He continued that the point was that he understood it, but a lot of the implementers do not. As far as they are concerned, there is only one backslash and that is in the zone file.

Patrik said that he thought that, in this case, it was even more important that the correct terminology be used when trying to explain the problem. He explained that one of the reasons why he thought that this was important was that the problem was actually much worse than Lawrence had described. He continued that when Lawrence said, "Just put a backslash there", it implied that all provisioning systems worked in the same way and that the same quoting was needed, which was not correct.

Jim Reid said that the backslash problem was primarily to do with provisioning in this context. He continued that BIND is talked about in many cases and this gets confused with DNS. He agreed with Patrick and continued that people had to be careful regarding delimiters and backslash, but the caution is not necessarily DNS; it may be a DNS implementation.

Patrik said that, as many attendees knew, he was one of the inventors of the thing that Lawrence was talking about. He continued that one of the things that he had noticed and one of the things that people might have seen in some of the recent RFCs is that he had been strongly encouraging the people who invent standards not to use the NAPTR record.

Patrick continued that the last slide of Lawrence's presentation was a question: "Should we do ENUM version two without NAPTRs while we don't have any real significant deployment?". He thought that this was an interesting thing to discuss.

Lawrence said that getting NAPTRs fixed would be good. He said that it could be done but that it would take a long time. He added that he thought that one of the reasons there was not a wide deployment was because it had taken so long to get the standards completed and even then the standards were unreadable.

Peter Koch said that Lawrence had been criticizing NAPTR for quite a while and there are operational reasons for this. He continued that he was not questioning Lawrence's motives but offered the observation that the reason that DDDS came into play was because someone found out that when the RFC for NAPTR was reissued, it was embedded into the DDDS scheme. He said that there was no correct way in terms of protocol to just use NAPTR and that it could only be used within these DDDS or within these DDDS schemes as ASNAPTR and UNAPTR.

Peter continued that what seemed to be missing and what seemed to make life harder for the programmers and implementers of these provisioning systems was that there was no generic DDDS library. If that would exist, things might be easier as it seems that people are running into this issue and implementing the details over again, which is the same as if someone would have to implement their own IP stack.

Lawrence said that he had mentioned that DDDS was over general. A general DDDS library would be too large. He added that he thought that the biggest problem would be the provisioning system.

Patrik said that he also didn't think it was possible to do a generic DDDS library. He said that some sort of application-specific things implemented for ENUM could be interesting. He asked Lawrence if he had been in contact with the Object Naming Service (ONS) to see whether his findings were similar to theirs and whether some of his suggestions are things that help them or hinder them.

Lawrence said that he had not contacted the ONS because he had been focused on the pure telecoms side of things.

Patrik asked if any attendees had worked with ONS. There was no response.

Jim said that he thought some more test cases could be produced. He added that although Patrik was on the back-end of the provisioning side of things, other things are also needed for clients. He asked whether there was anybody interested in trying to put together some sort of test bed to do testing on clients as well as testing on front-end or back-end provisioning systems.

Lawrence said that he thought that this was an extremely good idea. He asked whether there should be inter-op testing. The response was yes. He also asked whether torture testing, should be done. The response was that it definitely should.

E. ENUM Operations

E1: RIPE NCC ENUM Update
Anand Buddhev
http://www.ripe.net/ripe/meetings/ripe-58/content/presentations/tier0-enum-update.pdfhttp://www.ripe.net/ripe/meetings/ripe-58/content/presentations/tier0-enum-update.pdf

Carsten Schiefner said that the RIPE NCC used to have statistics on non-delegated country codes and asked if these statistics were still being collected.

Anand responded that he presented these statistics at the RIPE 57 Meeting in Dubai. He continued that this had been a one-off exercise rather than a continuous monitoring project and so there were no statistics available for this meeting. He continued that if this was something that attendees feel that the RIPE NCC should monitor on a continuous basis, he would consider it valuable input and make it an action item. He asked for a show of hands to see how many people were interested.

Niall said that the working group also exists on the working group's mailing list outside of the RIPE Meetings and so the question should also be asked there.

Carsten said that he would put the question to the mailing list about whether statistics on non-delegated country codes should be added to the RIPE NCC's report on a regular basis.

F. Short News

F1 enumdata.org Update
Niall O'Reilly


Carsten asked if those attendees who knew people responsible for the Tier-1 registry could ask them to send updates to the ENUM Working Group co-Chairs. He also asked that people also get in touch with the co-Chairs if they spotted something wrong or not entirely correct.

Niall added that there is a standard list of questions and asked that people please send the responses in plain text.

X: Interaction with Other Working Groups

AP WG Policy Proposal 2008-05

Y: A.O.B.

Denesh Bhabuta gave an update about UK ENUM.

There were no questions or comments on Denesh's presentation.

Niall pointed out that there was scope for talking about infrastructure ENUM in this working group as well, provided that people would come forward with related agenda items.

Z. Close

Summary of action items:

Action 58.1: Carsten Schiefner to put the question to the mailing list about whether statistics on non-delegated country codes should be added to the RIPE NCC's report on a regular basis.