RIPE 64

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

Wednesday, 18 April 2012, 14:00-15:30
Grand Union Hotel, Ljubljana, Slovenia
Co-chairs: Niall O'Reilly, Carsten Schiefner

A: Administrivia

Welcome, Scribe, Jabberwok, Microphone etiquette, Agenda

The ENUM WG Co-chair announced that items D.1 and D.2 would swap in the agenda.

B: Minutes of previous meeting

The minutes from RIPE 63 were approved.

C: Review Action List

ENUM-AP-63.1 - [OPEN] Denesh Bhabuta, Peter Szegedi: Contact ENUM operators‚ brainstorming among ENUM operators

ENUM-AP-63.2 - [DONE] Niall O'Reilly: Move the enumdata.org website

D: Main presentations

(Items D.1 and D.2 switch places on the agenda)

D.1 ENUM alternatives - Alexander Mayrhofer, enum.at

The presentation is available at:
https://ripe64.ripe.net/presentations/151-RIPE64-ENUM.pdf

Peter Koch, DENIC, referring to slide 6, asked why ENUM validation is a failure and if it didn't work as expected and if people got their numbers hijacked.

Alexander replied that it was a failure from the market perspective, it was too complex. It turned away most of the users. It worked where the carrier had the valid information.

Peter commented that that was only one perspective.

Alexander replied that that was his take on it.

Carsten Schiefner, ENUM WG Co-chair, asked if it was really the case that the validation part was complex and over engineered. He added that from his point of view, ISPs were not in a position to easily make money out of that technology. He said that it reminded him of turbo chargers and wondered if there's a consumer market for it. He added that he doesn't blame the delegation part but the inability to come up with attractive use case scenarios.

Alexander clarified that he meant to mix the technical and administrative things in the first slides.

Robert Schischka, nic.at, commented that Alexander was right about the products not being the problem but that they tried to make validation as market neutral as possible. He added that it is very nice technically, but no one took the chance at the market. Too much effort has been spent making something that doesn't meet the market requirements.

D.2 XConnect/DE-CIX NGN Peering Co-op - Wolfgang Tremmel, DE-CIX

The presentation is available at:
https://ripe64.ripe.net/presentations/93-e-wtklxc-RIPEGlobalVoix-003.pdf

Alexander Mayrhofer, enum.at, asked if they did other modifications on the ENUM query string to carry the source and destination phone.

Andrew Cooper, XConnect, (via Skype) replied that no, what's important is that the services know other information beyond the standard ENUM. We need to know the A number, the user agent and lots of other potential data we would get from a normal SIP header. We put a dot notation in front of the standard ENUM number. This is only internal and has the standard ENUM query format in reverse dot notation. SIP gives you all the information you would ever want, but is a much heavier protocol. He added that the question is really, is this where ENUM is going and is it going to give us enough information, or should we just go with SIP.

Carsten Schiefner asked how many service providers do they have on board already in Germany and what he sees for the foreseeable future.

Wolfgang replied that they are planning to roll out in July. They had a meeting with their first customers. Thirty were present and that's the size they'd like to start with. They are targeting 50-100 customers.

Carsten asked if it was internationally.

Andrew Cooper replied that XConnect has about 150 service providers connected for various services. He said that the tough question was how many numbers does that constitute. He said they have the North American number portability Database in their network and it's a lot of numbers.
He added that the industry itself is only just coming around to real true VoIP interconnect. It's still a tier 2 or 3 thing to do, but now you see many more tier 1. He said he thinks it'll grow very quickly. Eighty percent of voice is national, which is why they work with companies like DECIX to provide high quality nationally. So there are two markets. The IPX providers are generally focused on the international market and they look at the national markets and linking them.

Carsten asked if there are competitors or alternatives in Germany and commented that large telcos may have thought of their own solution.

Andrew replied that PSTN is still the interconnect that they want to do. So they're competing with the transit carriers like DTAG. The start is to move away from that model and support a true multimedia interconnect point and support video and support all those kinds of things. Andrew said he didn't think anyone is really looking at this at that level. These transit networks make money out of transit for voice calls. But it is almost impossible to do today, so interconnection will be much more of a commodity with everyone realising they have to do it.

Peter Koch, DENIC, said he was confused about the use of the term "ENUM technology" because what Andrew called enhancement of the standard was kind of an extension of putting stuff in the DNS. He asked if anyone knew other enhancement or additions to the ENUM protocol that are "ENUM tech" but incompatible on the wire and what that means on this forum. He asked how long they would continue to call this ENUM.

Andrew replied that the question is "What does ENUM mean". For them, it's a means to enable customers to query their database and registry. Theoretically they can use SIP, ARQ. They are looking at ENUM as a protocol and a commonly supported method between networks. He asked whether ENUM is the structural way that DNS works. He said that no, from their perspective, it's not and they are separate. The ENUM standard in terms of protocols, data structure, as a protocol standard between two network elements, versus the bigger picture of saying ENUM will be zoned like DNS and delegated at the country code level.

Peter commented that the term ENUM is used as a higher-level basis for IP or Internet technology as a lookup for the routing of the phone call. He said the distinction between the lookup part and the routing part is worrisome.

Niall O'Reilly, ENUM WG Co-chair, commented that there is another taxonomy to describe whether you're violating the concept of ENUM, meaning that you're using the DNS in a way whereby you take the parameters you might be negotiating during SIP and encode those in the DNS namespace and use that to pre-qualify the call to decide if it is worth bothering with the more complex SIP action if the call is going to be accepted.

Andrew replied that yes, the registry is a term that sits between policy and routing. He said that what they are doing is querying that to find out what's allowed or not. He said that they have all the information you could possible want about what route to return if you query via SIP but if you query using ENUM, who is originating service and what's the B number? They don't know the A number or user agent or anything. They are using the protocol and extending it to make a richer policy and routing decisions.

E: ENUM Operations

Tier-0 Report‚ Ann Barcomb, RIPE NCC

The presentation is available at:
https://ripe64.ripe.net/presentations/26-RIPE64_Enum_Update.pdf

There were no questions.

F: Short News

F.1 enumdata.org Update, Niall O'Reilly

The presentation is available at:
https://ripe64.ripe.net/presentations/28-RIPE-64-ENUM-Status.pdf

Action item "ENUM-AP-63.2: Niall O'Reilly: Move the enumdata.org website" has been completed.

F.2 Contact ENUM operators‚ Denesh Bhabuta

Action item "ENUM-AP-63.1: Denesh Bhabuta, Peter Szegedi: Contact ENUM operators‚ brainstorming among ENUM operators

Denesh said that he was getting frustrated with the ENUM updates being the same and that it was apparent nothing was happening in end-user ENUM. He said they had a panel to discuss this. They emailed 19 operators with some questions and only got three responses so no conclusions can be made. He said he hoped for more conclusions to present at RIPE 65.

Niall asked if he wanted to keep this action item open.

Denesh said yes.

F.3 News item

Niall informed the audience that the Canadian regulator brought out a regulation on interconnect that all who offer interconnect over IP must also offer it to anybody else who is interested.

G: Discussion on plenary presentation

X: Interaction with other working groups

Y: AOB

There were no AOBs.

Z: Close

No new action items created, but one still open.

Summary of open action items:

ENUM-AP-63.1: Denesh Bhabuta, Peter Szegedi: Contact ENUM operators, brainstorming among ENUM operators