PBT is just another protocol. Yet somehow, in the service provider telecommunications world, protocols unleash exorbitant amounts of argument and passion which tend to then get mired into technical detail.
 |
|
|
|
|
Is there room for PBT in the networks of the future? Absolutely. It just won't take over the world.
Eve Griliches
Program Director, Telecommunications Equipment, IDC
|
|
|
|
|
|
|
|
|
|
The bad part of this is as the technical details get, well, more technical, the observer gets lost. What did they start with? Lower cost, simplicity? All that sounded good. The fact is, no protocol is easy to deploy (MPLS), no protocol comes feature-rich on day one, and no protocol is the answer to all questions.
The service provider side
But that doesn't answer where are we today. We have been introduced to PBT, listened to the hype, seen BT announce it's an advocate and will deploy it, and other providers say they are "interested." But now with management changes at BT, PBT is no longer the focus, and the company will continue with its MPLS deployment.
Although PBT was not implemented immediately, this does not mean BT will never deploy PBT in its network, just that it won't be deployed right now. Verizon Business units have also hyped PBT, but we've come to learn who really runs the Verizon network, and it's the core team in Texas that make the decisions. Those folks never wanted control planes to move to the "management" area that PBT requires.
PBT vendor support
So that's the service provider side; what about vendors that support PBT? Has Nortel or its "ecosystem" benefited from PBT coming to market? From the Carrier Ethernet revenues IDC tracks, it is not apparent that revenue has increased in Carrier Ethernet products for Nortel, NSN (Seabridge unit) or Extreme Networks. So clearly, new protocols do not help sell specific hardware platforms.
But now that we've established this, is there room for PBT in the networks of the future? Absolutely. It just won't take over the world. PBT is well suited for transport network providers that want to migrate from SONET/SDH to Ethernet for lower cost points but still have the control and point-to-point circuit-based management they've always deployed.
The real question is, do you build an Ethernet services network over an MPLS infrastructure that exists today or spend some marginal dollars on a new network addressed specifically for Ethernet? The answer isn't entirely clear or alarmingly simple, mainly because no one has the luxury of building a new network today. All networks will have to migrate from where they are to cost-effectively deploy new services.
So while PBT had some over-hyping in the market, at the moment, we believe it's also getting an over-bashing. Protocols simply take years to become feature rich, fully tested and appreciated. And there is room for multiple protocols in the networks based on the profiles of the providers that deploy them. So our advice is don't swing with the pendulums too much. What is over-hyped today may be widely deployed in the next few years. And what takes years to deploy may actually become so feature rich that it slows migration to newer protocols. This is definitely a 'stay tuned' to area.
About the author: Eve Griliches is a Program Director within IDC's Telecommunications Equipment group. She provides in-depth insight and analysis on service provider routers and switches as well as the optical networking market. Ms. Griliches also provides critical business intelligence on emerging technology trends and their impact on the overall telecom market space. Ms. Griliches joined IDC in 2005 after 10 years in product management for a number of network equipment vendors. Most recently she worked at Marconi, as well as PhotonEx a start-up in the 40Gb/s optical long haul market. Prior to that, Ms. Griliches held positions at Wellfleet/Bay Networks/Nortel Networks in the Carrier Routing Division with strong interworking with Nortel's Optical groups. Additionally, she spent 4 years at Thinking Machines Corporation, one of the first parallel processing supercomputer companies.
