Provider backbone transport technology vs. multi-protocol label switching has been one of the hottest technology debates of the last year or so. PBT arrived on the carrier Ethernet scene promising less expense and more efficient point-to-point connections than the MPLS evolution could muster. It rose to popularity the way new technologies often do--it had at least one vendor (Nortel Networks) willing to make a case for it and at least one major carrier (BT) willing to bet on it (BT chose Nortel and Siemens last year for an eventual PBT deployment. More carriers started talking about PBT and testing it, and pretty soon, a whole ecosystem had arisen.
Now, BT says it will not be deploying PBT to the extent originally envisioned, instead favoring PBT nemesis MPLS. That decision is being seen as a gut-punch to Nortel, a company for which PBT had appeared to be one possible engine of recovery after a horrible period of financial upheaval. If BT's decision to play it low-key with PBT is some sort of death knell for the technology, Nortel does stand to lose big, but it is not the only one. What about all the companies that have been developing PBT control planes and testing suites and components? What about all the vendors that got into the business because Nortel and Siemens (now Nokia Siemens Networks) got that first nod from BT? And what of the ongoing PBT standardization process (in which PBT goes by the more cumbersome acronym PBB-TE, for provider backbone bridging-traffic engineering)?
Could one decision by one carrier have such far-reaching consequences? Such a decision of one of the most significant international carriers certainly cast a shadow over PBT for now, but that shadow could lift, and here's why: PBT was conceived as a solution to much of the Ethernet management complexity that carriers did not want to deal with as they looked to expand their Ethernet capabilities and offer more point-to-point Ethernet transport services. PBT can still do that, and standardization could further broaden its appeal to carriers who like to wait for a clear standard to be in place.
Early on, some industry observers said PBT and MPLS were likely to both play roles in evolving carrier Ethernet networks, but our tendency to debate and argue and grant one side an edge may have pushed such observations into the background. PBT solves problems carriers wanted solved. For that reason alone it has a future, even if the future suddenly does not seem as immediate as first thought. It may even still have a bright future with BT. The U.K. telco is not shutting out PBT, and it may still find an increasing need for it alongside MPLS and supporting different connection needs than MPLS.
PBT will not die with on the whims of one carrier, but what it most needs now are other carriers to continue evaluating it, and provide the industry with a roadmap for when, where and how it will be used.
- read this story about the effect on Nortel at Light Reading