Mike, thanks for bringing that post to the board. I agree with the author on most of his points, especially from an historical (meaning up until right now, for most situations) perspective. And above all, I like the way this author characterizes the marketecturists.
The author tends to generalize a bit too much, IMO, as he goes into his supporting paragraphs, and in so doing leaves himself open to a growing number of potential corrections, depending on the particular city or town or rural area that one finds themselves in. And, as we look over the horizon into future, the number of dissimilarities, hence corrections, possibly increases, again based on where you are and how much competition exists in that locale.
Especially in the last-mile/access/edge portions of "the network" where some of the specialized optical (and soon wireless) platforms going in today for the first time will not now, nor will they ever, accept the same types of SONET boxes that were previously deployed deeper into the network, i.e., in its core, at an earlier time. IMO, he's focused on those aspects which do not change for those areas where stagnation will prevail, but does not focus hardly enough on those areas where disruptions to the networking world's status quo are taking place.
[[Begin Sidebar. A case in point: I know of a large commercial enterprise on the West Coast for whom I once designed and RFP'ed a private SONET network. In order to support IP traffic in the metro, as well as the wide area, they used ATM at 155 Mb/s and then 622 Mb/s (OC-3c, and then OC-12c, respectively), and sent their IP over ATM over SONET. That was only four years ago. Yes, this is a very large software company.
They did this for both their WAN and MAN stuff. By MAN, I mean their intra-city/virtual campus stuff that comprised sixteen building spread out over a twenty-five mile radius.
Today they are installing a 4-strand (multi-wavelengths per strand) private fiber network between each of their sixteen buildings. On these they will be using a new flavor of SONET (not something that existed before, for its ability to control newly RFC'ed (which stands for request for comment documents which are proposals for new standards) Internet Engineering Task Force (IETF) protocols in addition to ANSI-ITU SONET/SDH. But more to the point, they will also be pushing GbEthernet over NON-sonetized fiber now, and sometime within the next eighteen months they are already planning to pilot 10 Gb Ethernet which when stabilized will be used for the majority of their network backbone connections. End Sidebar.]
On the other hand, for most of the established LECs, and for many inter-exchange carriers and their resellers and other forms of parasitic entities, what he has stated is precisely what will take place.
It was once very easy to make broad brush statements when the network was a virtual monolith, where every carrier, no matter who it was, was using the same architectural configurations made by the same vendors, which in sum came to make up the public switched telephone network (PSTN) and its supporting high capacity undercarriage which we've used for data and now Internet, as we've grown to know it.
But today, there are too many flavors of networking, to many different protocols, too many vendors whose boxes may or may not interoperate with one another to choose from, and the number keeps growing. It's therefore very difficult to make general statements, which cover all of the bases, in such an environment.
He's referring to the classical migration strategy, which, like he says, has been in place ever since A. G. Bell spilt battery acid on his crotch. But he doesn't readily acknowledge the new set of demand dynamics that are being fostered by residential broadband, the non-Bell data clecs bringing "next gen" services to small businesses, the fiber barons (esp. the metro guys like MFN and Telergy) who are coming onto the scene now with no such older baggage to migrate from and to.
Then there are the metro resellers and integrators of optical-based facilities who are bringing enterprises onto platforms based exclusively on lambda switching, IP routing, and a host of new networking techniques which are too exotic to get into here, which don't even remotely resemble the telco-subscriber networking fabrics of the past.
While reading the post I kept waiting for the author to begin speaking about the resource demand shifts that are now taking place, the ones which are coming from the other end of the pipes, and then he finally [in the last paragraph] acknowledged the new influences being brought about by the "broadband" (I still have to cough every time I use that term) technologies of dsl and cable modem.
All in all he quite aptly portrayed how legacy carrier hand-me-downs take place, both now and in the past. But one has to wonder how these migration strategies will be altered in the future when the sucking sounds begin to reverse directions at the edge, or when they change dynamically and unpredictably with tsunami-like force.
--------
You know what? In re-reading the post again, I've come to the conclusion that everything I've stated above could have been stated by this same author if he knew that he was going to be critiqued. Just a gut feel. It comes from learning how to decipher concepts which were only half stated, and reading between the lines when I know that someone is trying very hard NOT to show off all that he or she knows. smile
FAC |
