RPR Gurus Set Fall Deadline

The newest thing about the proposed standard for Resilient Packet Ring (RPR) technology is: The new stuff must stop.

The Institute of Electrical and Electronics Engineers Inc. (IEEE) 802.17 RPR Working Group says that as of the end of September, it won't allow any more significant technical changes in its draft of the RPR technology standard. The move is meant to speed up the process of getting RPR from draft stage to full IEEE approval by Fall 2003, according to Bob Love, vice chairman of the IEEE's 802.17 working group.

RPR is a technology that aims to enable public telephone networks to carry data traffic more efficiently and provide the redundancy common with exiting Sonet technology. It also helps connect Ethernet-based local area networks to the Sonet-dominated public telephone networks.

The technology gives Sonet, designed for handling voice traffic and point-to-point connections, some Ethernet-like characteristics, such as being able to add new network nodes quickly and use bandwidth more efficiently. It also gives Ethernet, designed for moving packet-based data traffic, the Sonet-like characteristics of being able to operate on packet rings and being able to recover from a cut connection within 50 milliseconds.

The RPR Working Group is made up of more than 400 participants from more than 100 companies, according to the Resilient Packet Ring Alliance. Among the members are several startups -- including Corrigent Systems Inc. and Lantern Communications Inc. -- that have bet some or all of their business on RPR's eventual acceptance.

Rather than incorporate new features to the RPR working draft this time around, Love says that, for the sake of schedule, such features will have to wait until a later version of the technology is introduced. "We did not want to be delayed on something that is an extra, but not critical, bell and whistle," says Love.

Considering the history of RPR, this is a noteworthy accomplishment. In late 2001, two main groups -- one led by Cisco Systems Inc. (Nasdaq: CSCO), the other by Nortel Networks Corp. (NYSE/Toronto: NT) -- were at a standstill on which version of RPR would survive to become the industry standard. In January 2002, however, the two camps submitted a joint proposal, and that became the draft that the RPR Working Group has since been tweaking (see RPR Moves Forward).

The September cutoff for technical changes only applies to new features and not fixes of the current draft, Working Group members say. Likewise, there's no telling what future changes might be introduced before the IEEE has declared RPR an official standard.

After the RPR Working Group's September meeting, and after subsequent Working Group comments on the draft, the technology draft must pass muster with a larger industry group, via a sponsor balloting process, before it is submitted for final approval as a technology standard.

In the meantime, equipment vendors are pushing earlier, proprietary flavors of RPR, the most dominant of which is said to be Cisco Systems’ Spatial Reuse Protocol (SRP)/Dynamic Packet Transport (DPT).

By eliminating many of a network's point-to-point Sonet connections, RPR in theory will reduce the need for adding new next-generation Sonet add/drop multiplexers. But the ring-based RPR may also run afoul of some Ethernet equipment vendors, which would prefer that carriers use a mesh network.

"The real point about RPR is: Why bother?" says Andrew Knott, VP of marketing and customer service at White Rock Networks. "Carriers generally only absorb new technology in large quantities if two things are true: It has to be noticeably cheaper and more manageable than what they're doing today."

RPR proponents, however, note that RPR does simplify networks by incorporating the best characteristics of Sonet and Ethernet. "We think of RPR as a completing, rather than a competing, technology," says Love.

In an upcoming Webinar, Light Reading will discuss RPR technology and equipment in more detail. For more info, visit: http://www.lightreading.com/webinars.

— Phil Harvey, Senior Editor, Light Reading
http://www.lightreading.com Want to know more? The big cheeses of the optical networking industry will be discussing this very topic at Opticon 2002, Light Reading’s annual conference, being held in San Jose, California, August 19-22. Check it out at Opticon 2002.

Register now and save $500 off the registration fee. Just use the VIP Code C2PT1LHT on your registration form, and deduct $500 from the published conference fee. It's that simple!

Bumper_car 12/4/2012 | 10:04:53 PM
re: RPR Gurus Set Fall Deadline For those that are looking at RPR, there is a protocol standard from ITU called G.gfp (Generic Framing Procedure) that also supports transport of diverse data protocols over rings. This is "mapping" protocol for the successful and widely deployed SDH/SONET transmission systems. Where does RPR sit in this?
gbennett 12/4/2012 | 10:04:47 PM
re: RPR Gurus Set Fall Deadline Hi Bumper_car,

GFP is now standardised as G.7041 and, as the name suggests, it's just a framing procedure.

In contrast, RPR is essentially a new MAC layer for optical networks including SONET and SDH rings. Note that there are also plans for RPR to be used on both dark fibre and dark wavelength networks.

Normally SONET and SDH are TDM muxing schemes. This means that the total amount of access capacity cannot exceed the capacity of the trunk. For voice and circuit emulation this is just fine as it delivers a high level of QoS that these applications need. But data is very bursty and it's difficult to provision the TDM timeslot efficiently. So, as a MAC layer RPR allows a SONET/SDH ring to be over subscribed for data traffic. In other words it allows some portion of the ring's bandwidth to be statistically multiplexed, and therefore used more efficiently than static timeslot assignment for bursty data.

RPR may also allow some amount of TDM-like traffic to pass across the ring with a guaranteed bandwidth, delay and jitter level. This aspect of RPR is the one under most scrutiny since the "compromise proposal" was agreed by IEEE 802.17. Fans of RPR say that there's still enough traffic management in the proposal to allow the efficient sharing of voice, circuit emulation and LAN data. Opponents of RPR feel that the compromise proposal was too "watered down" by Cisco, and is nothing more than DPT with an IEEE stamp of approval. The reality is that it's too early to say either way as it will be months before real RPR implementations hit the streets. Anything out there today that claims to be RPR is Marketing At Work.

One other aspect of the RPR MAC is that it offers a similar, ring-based protection scheme to SONET/SDH, but with spatial reuse. In short this can mean the end to the 50% "idle bandwidth problem" in 2-fibre SONET/SDH rings.

Hope this helps.

stuartb 12/4/2012 | 10:04:33 PM
re: RPR Gurus Set Fall Deadline I think the collective yawn in the telecom industry momentarily reduced the amount of breathable air here on Earth.

RPR was envisioned at the height of the bubble. There are no compelling business reasons for a carrier to deploy RPR, and it will continue to whither and die.

optigirl 12/4/2012 | 10:04:33 PM
re: RPR Gurus Set Fall Deadline Wonder what Mr. Jay, RPR Will Rule The World" Shuler or Luminous will reply with...

Sign In