Ethernet equipment

Analyst: PBT’s Not Dead Yet

PBT may have lost the support of its loudest carrier supporter, but the decision by BT Group plc (NYSE: BT; London: BTA) to favor MPLS in its 21CN rollout isn't the death blow to the emerging carrier Ethernet technology that some might think. (See PBT Sidelined at BT and A Guide to PBT/PBB-TE.)

So says Heavy Reading's Ethernet specialist Stan Hubbard. He noted last week, when Light Reading broke the news of BT's decision, that the British operator's U-turn is "a temporary setback associated with finding the right mix of PBT and MPLS in the metro network."

Now Hubbard adds that BT's decision is just part of a more widespread decision-making process around PBT that will likely see other major carriers coming to the same conclusion as the U.K. incumbent, but also result in some positives for PBT's vendor supporters, led by Nortel Networks Ltd. . (See Nortel: There's More to PBT Than BT.)

Hubbard says 2007 was "obviously a banner year for the marketing of PBT/PBB-TE after BT announced that it would be using the technology in 21CN, and a number of companies joined the Nortel-led carrier Ethernet ecosystem." (See Nortel, Siemens Win PBT Deals at BT, BT Goes Live With PBT, Nortel Touts Ecosystem, TPack Talks Ethernet, Gridpoint Joins Ethernet Ecosystem, Soapstone Intros PNC for PBT, Extreme Joins Ecosystem, and Nortel Pushes PBT Pact.)

"We really looked to 2008 for PBT/PBB-TE pioneers such as Nortel, Extreme Networks Inc. (Nasdaq: EXTR), and Nokia Networks to begin delivering on the promise, but the news has been mixed so far when it comes to PBB-TE trial activity, announced customer wins, revenues, and flexibility in supporting multiple service types," adds Hubbard.

"Research by Heavy Reading and other analysts indicates at least a dozen Tier 1 operators have been, or soon will be, evaluating PBB-TE solutions. On the upside, Nortel has signaled that it expects to turn some RFPs/trials into new Tier 1 wins in the second half of the year," continues Hubbard. "Also, other vendors recently told me that they have received positive feedback from some large operators who have been examining PBB-TE."

So there's hope yet for the PBT fraternity -- but it's not all positive news.

"On the downside, we have BT's plans to scale back the role for PBT in its network. In addition, my own discussions in mid-April with three Tier 1 operators -- two in Europe and one based in the U.S. -- indicated that they had examined PBB-TE and were not likely to embrace it at a corporate level any time soon," adds the analyst. So BT is definitely not alone in finding PBT wanting.

Hubbard continues (he has big lungs): "Beyond BT and China Netcom Corp. Ltd. (NYSE: CN; Hong Kong: 0906), essentially all of the publicly announced wins have involved Tier 2 and Tier 3 operators. That’s where Nortel has gotten traction, and that’s the key market Extreme is going after." (See Nortel Wins in China.)

And at least that business has been generating some revenues for the vendors. "Nortel has generated PBT revenues for several quarters. Extreme recently stated it is expecting PBB-TE revenue before the end of 2008. But Nokia Siemens, which now has its own Layer 2 MPLS story following the Atrica acquisition, hasn’t said much publicly about PBB-TE sales."

But Hubbard notes that the issue of multipoint connectivity is likely to be a major issue when it comes to carriers' decisions about PBT deployments.

"One of the biggest concerns raised about PBB-TE is that it only provides point-to-point connectivity. That works fine for dedicated Ethernet Private Line and mobile backhaul transport -- two key applications where Nortel has found traction -- but operators also are looking for point-to-multipoint and multipoint-to-multipoint solutions," notes Hubbard. And indeed that's exactly the reason BT has given for swinging toward an MPLS-focused strategy.

But the PBT camp is, of course, aware of this and trying to come up with a multipoint version of PBT. "Nortel and other vendors recognize this as an issue, and you’ve seen from Nortel that they plan to address this in the second half of 2008," notes Hubbard, referring to Nortel's plans to unveil a new carrier Ethernet switch, dubbed MetroNext. (See Nortel Preps New PBT Switch.)

— Ray Le Maistre, International News Editor, Light Reading

mr zippy 12/5/2012 | 3:39:43 PM
re: Analyst: PBT’s Not Dead Yet I think PBT, or a similar "carrier class" ethernet based technology, is inevitable.

We've seen multi-protocol layer 3 networks (IPX, Appletalk, IP etc.) converge to the of single protocol IP. This is despite IP being both an older and less capable technology that the networking protocols it replaced. I think the main drivers were access to the Internet and then the simplicity of running a single protocol network, with increased simplicity reducing both capex and opex costs.

MPLS provides a more unified interface to different layer 2s, making the layer 2 interface presented to the upper layers more consistent, resulting in it being the so-called "layer 2 and a half." For example, MPLS hides the differences between the sizes of the Frame relay and ATM VC fields, or adds a VC field capability to protocols like ethernet or PPP that don't have one.

We're now seeing a similar single networking protocol convergence occurring at layer 2. Everything is either converging to ethernet or an ethernet shim layer is being placed over the top of it. The drivers for that are the commodity cost of ethernet and again, the simplicity of running a single protocol network.

Once layer 2 is converged to a single network protocol, some of the capabilities of MPLS used to hide the differences between the different lower layer 2's aren't necessary. Alternatively, when adding capabilities to ethernet, such as P2P TE paths, they can be designed to only suit ethernet. MPLS's unnecessary "layer 2 adaption" features will end up being dropped, and other useful features of ethernet, such as globally unique node ids, rather than per-hop unique ids (i.e. VCs), can be taken advantage of.
Sign In