By CAROLYN MATHAS
Although controversy, dissension, misrepresentations, and delays have been reported for both resilient packet ring (RPR) and metro Ethernet standardization efforts, both camps are reporting substantial headway on their paths to becoming viable transport standards after recent meetings.
Since its inception in December 2000 after a full year as a study group, the RPR 802.17 Working Group defines RPR technology as an alternative to SONET transport for packet networks. RPR will deliver resiliency, fast protection and restoration, and performance monitoring comparable to those of SONET networks, say proponents.
During the mid-November meeting held in Austin, TX, the working group claimed the following milestones:
- An agreement on the outline for the RPR draft standard, which includes all major sections of the standard that will be filled in as consensus on each area is reached.
- Reached consensus and combined technical proposals regarding the physical layer and management portions of the draft standard, which effectively completes both sections.
- Three proposals offered to merge ideas to eliminate those segments of proposals that do not exhibit backward-compatibility.
"We made significant advancement toward converging technical proposals," explains Bob Love, chair of the RPR Alliance and vice chair of the IEEE 802.17 Working Group, ex plains. "I am confident that we will continue to coalesce the remaining technical proposals by our January meeting and publish a first draft shortly thereafter."
The next meeting of the IEEE 802.17 Working Group is scheduled for Jan. 14-17 in Orlando, FL.
According to Nigel Cole, vice president, business development at Corrigent Systems (San Francisco) and member of the RPR Alliance, "RPR set out to address the problems with Ethernet, which has been touted as the new low-cost transport infrastructure-much cheaper than SONET. The problem is that it was never designed for the carrier space."
John Hawkins, treasurer of the RPR Alliance and senior marketing manager for Nortel Networks (Ottawa, Ontario) agrees. "RPR focuses on what we believe to be the key problems and requirements of the RBOCs today in the metropolitan-area networks," he explains. "Historically, their revenue is based on the sale of voice services. They are now faced with having to support a tremendous number of data bits and transport them across the network in a way to meet their customers' requirements, including SONET reliability, 'five-nines' availability, and the ability to reconfigure within 50 msec. We see our basic requirements are the support of the ring topology, the support of voice with the same reliability and quality of service available today, and the ability to add massive amounts of data traffic."
The economic downturn shouldn't slow the requirements that RPR is looking to meet, according to Raj Sharma, vice-chair of North American affairs for the RPR Alliance and director of product management at Luminous Networks (Cupertino, CA). "I think the economic downturn has made a very strong case for RPR. One of the things that carriers are thinking today is the fact that they need to make more money off the assets in terms of fiber. RPR fundamentally introduces the concept of packet switching as opposed to circuit switching, which gives them an opportunity to be more profitable with the private-line services in demand lately," he asserts.
Hawkins further states that RPR can actually help carriers mine some of the gold they have in the ground. "It has really put pressure on vendors to go to a standard solution as opposed to continue using proprietary solutions. What we're hearing loud and clear is that the market wants standardization," he says.
Love acknowledges that some people say, "'It's not Ethernet'-and it isn't." However, he explains that if the purpose is to transmit Ethernet, RPR interfaces with the enterprise to seamlessly and effectively transport Ethernet data. The data is appropriately packaged within an RPR media access controller (MAC) and, upon reaching its destination, is unwrapped and delivered as Ethernet frames. "It is the lack of understanding that this will be part of the Ethernet structure that is one of the largest misrepresentations regarding RPR today," says Love.
Meanwhile, the 50-member Metro Ethernet Forum (MEF) reports porgress in its standardization efforts. Nan Chen, president of the forum and director of product marketing at Atrica (Santa Clara, CA), says the organization works to accelerate adoption of optical Ethernet in metro networks worldwide by defining carrier-class Ethernet-based metro transport technologies and Ethernet services. As its second priority, the forum also intends to define work to be done by other bodies such as IEEE 802.17.
"Ethernet services," according to Chen, "can be carried by Ethernet transport or by RPR, so it's transport-independent. No other standards body is defining Ethernet services, so we are the technical standards body that will do so. From the Ethernet transport perspective the issues of protection, resiliency, TDM support, quality of service, all of these transport issues need to be resolved."
At the forum's October technical committee meeting, the process of creating a new set of technical specifications was undertaken. As a result of the meeting:
- A protection model and implementation framework addressing 50-msec resiliency in the event link or node failure was developed.
- An initial Ethernet services model with definitions, profiles, and templates also was developed. Based on this model, service offerings are being defined and service-level-specification (SLS) frameworks created.
- An initial proposal for a quality-of-service framework was completed.
- An initial Metro Ethernet Reference Model was developed to enable a common "baseline" from which other technical subcommittees can work.
- A management subcommittee was formally organized and its charter adopted.
The third meeting of the MEF technical committee, along with the MEF plenary and marketing committee meetings, will be held Feb. 5-8 in San Diego to review various drafts developed by subcommittees, entertain additional proposals, and endorse the first drafts of MEF technical specifications.
Chen estimates that because the MEF is leveraging a substantial body of work, the standards process will take only a year.
As to the competitive aspects of RPR and optical Ethernet, Chen asserts, "The Metro Ethernet Forum does not compete with any other industry organizations such as the RPRA or 10-Gigabit Ethernet Alliance; indeed, many of the forum's members are also RPRA or 10GEA members or are members of other strategic alliances."
Chen says that recent reports of dissension stem from individual rather than organizational sources. Several members of MEF also belong to the RPR Alliance and 802.17 Working Group. "Although they are different folks, they come from the same companies. Both camps want to see their solution successful. Cisco, Nortel, and other companies really want to remain technology-agnostic. We're all working on our respective technologies and will let the market decide," says Chen.
"If you look at what's successful, it's not because it has a better technology, it's because it's cheaper, good-enough technology. It won't necessarily be RPR or Ethernet. It could be both. There are applications for both," Chen adds.