DUNDi Aims to Open VoIP Peering

By Sean Michael Kerner | Jan 4, 2006 | Print this Page
http://www.enterprisenetworkingplanet.com/unified_communications/DUNDi-Aims-to-Open-VoIP-Peering-3575106.htm

As the world's telecom networks move away from the plain old telephone service model to VoIP, the need for a new form of directory and interconnection is becoming increasingly apparent. Some believe the next generation directory to be ENUM (definition), others such as open-source revolutionary Mark Spencer see the future to be a tool called DUNDi.

DUNDi is an acronym for Distributed Universal Number Discovery and is by definition a "a peer-to-peer system for locating Internet gateways to telephony services. DUNDi is fully distributed, with no centralized authority whatsoever."

DUNDi is also very much the technical implementation of the open source ideal: many contributors, all with the opportunity to benefit and operating without a monopolists control. The open source nature of DUNDi is hardly an accident. DUNDi's creator, Mark Spencer, is, after all, the man who created the popular Asterisk IP PBX, of which version 1.2 was recently released.

Though DUNDi appears to many to offer great promise, it has its share of detractors and naysayers as well.

What is DUNDi?

Like ENUM, DUNDi is a system for number discovery that is mapped to an IP address. Unlike ENUM, with DUNDi there is no central repository for directory data. Instead, what occurs is that all DUNDi federated servers participate in a trust relationship whereby each node is connected to at least one other node in the network. The claimed benefit is that there is no central point of failure and there is also no monopolist control.

DUNDi can be used in an enterprise environment to provide greater redundancy and to connect PBXs, as well as for wider usage across the E.164 number space (definition).

How successful has DUNDi been in challenging ENUM in the year since its debut? The answer depends on your frame of reference. Spencer explained that determining DUNDi's current adoption success depends on whether you look at the enterprise or all of the E.164 number space.

"In the enterprise it's definitely catching on as the way to do clustering," Spencer told VoIPplanet.com. DUNDi has also been included in the Asterisk 1.2 IP PBX release, which Spencer expects will help adoption as well.

In the E.164 number space, Spencer explained that rather than trying to build an entire network with DUNDi from the ground up, he is now thinking of DUNDi as a way to connect people that are already providing services, enabling them to make their services interoperate without having to expose customer data.

"Instead of having this giant land grab that we've got right now in the ENUM space with everybody fighting over who is going to be the monopoly who owns that business, you can create a system where people can simply peer with one another," Spencer said.

DUNDi for peering
The VoIP peering market is one that is currently undergoing rapid change and growth. One of the major players in the space is Stealth Communications' Voice Peering Fabric (VPF). Stealth Communications president Shrihari Pandit president and CEO commented that DUNDi is appealing to enterprise organizations that deploy Asterisk and wish to interconnect them in a plug-and-play fashion. According to Pandit, it's still questionable if it's suited for carrier-to-carrier peering, based on how DUNDi functions.

"ENUM is a simpler and more realiable mechanism for handling peering between various organizations—carrier-to-carrier, carrier-to-enterprise, and enterprise-to-enterprise," said Pandit. "A single query into a ENUM registry will result in the exact location where the call can terminate within a few milliseconds."

That's not to say that the VPF wouldn't support DUNDi. Pandit explained that the VPF can support any routing protocols including DUNDi if the members wish to run it across the peering fabric.

"Today DUNDi is only supported in Asterisk," Pandit said. "In order for DUNDi to gain adoption, DUNDi will need to be implemented within session border controllers, soft-switches, and other VoIP gateways."

Barriers
Info-Tech Research Analyst Carmi Levy isn't overly optimistic about DUNDi's prospects.

"To be blunt, DUNDi will not succeed as a commercially available product that revolutionizes VoIP-based peering," Levy said. "It is certainly compelling technology that offers enterprises rolling out IP-based telephony a highly redundant option for peering services. But it has thus far failed to garner sufficient traction from the more mainstream vendors who would effectively deliver it to a corporate audience."

In fact among the vendors queried by VoIPplanet.com for this article only Nortel responded. And the response wasn't particularly favorable either. "Nortel has no current plans to support DUNDI in any of its IP PBX products/solutions," Nortel's spokesperson said.

Info-Tech's Levy sees greater potential for DUNDi in a disaster recovery scenario. Levy noted that a DUNDi-enabled Asterisk solution would facilitate easy replication of dial plans and full VoIP-based functionality much more so than current centralized ENUM-based peering. As an open-source alternative to commercial and centralized peering services, it is a relatively low-cost solution for IT managers looking for increased redundancy and failover.

"It doesn't mean that DUNDi won't eventually gain momentum in the telephony market," Levy said. "It simply means the road to market success will not be an easy one."

Spencer himself is clearly aware of the daunting task that faces DUNDi. In his view, would-be ENUM monopolists like VeriSign and Neustar, as opposed to any technical challenge, represent DUNDi's biggest barrier to adoption

"I don't think that companies like Neustar or VeriSign are likely to sign onto DUNDi anytime soon because they all think they've got a shot at being the monopolist," Spencer said. "So they are out there trying to win that war."

Spencer added that the winners with DUNDi are the smaller ITSP providers and the market as a whole as opposed to any one entity. Spencer's company Digium doesn't have a business model for DUNDi either as he believes that DUNDi is a technology that is in the best interest of the market and that all users can and should benefit from it.

"On the other hand you've got all the companies fighting to be a monopoly that have really big dollar signs in their eyes about how much money they'll be able to make with this," Spencer said. "So those people have a very strong incentive to make sure that DUNDi doesn't work."