Dan, just a few points I'd like to make on the upstream commotion, which is not necessarily directed to you, although I'd enjoy hearing your views in reply, as well.
There has been quite a bit of furor concerning the manner in which ATHM invoked its limitations in the upstream. On Advantage, indeed. They call it an enhancement.
As much as I don't like to admit it, this happens to be a prudent measure that most network admins must take when they are going through growing periods. Translation: Always. Unless, of course, they so outrageously overbuild their plant at the outset at costs so great that they can never recover them in a timely manner. But this is not the way the big leaguers do it. They grow incrementally, to meet visible and identifiable needs, even if those demands are in arrears. Better to have that kind of problem, they figure, than to build on the gamble and get no takers.
These are some of the factors behind what we now have in the upstream controversy, and I stongly suspect we will also see in the downstream, before long. Otherwise, we wouldn't have seen so much recent press from T about how they plan to resegment fiber and coax down to 50 to 75 homes passed, as well as reworking the manner in which they distribute power, in the SLC trials.
No network is infinite in its capacity, despite the improvements afforded by fiber. Available bandwidth is not a measure of the medium itself (in this case HFC), rather, it is a function of the port sizing (both phsycial and virtual) and the allocations that are made in spectrum (the frequency plan) that makes up the distribution scheme that are realized at the I/O levels. Said any other way, the end result is the same: The determinants are the provisioning of the spectrum allocation and the sizing of the ports, not the raw carrying potential of the medium (fiber).
The only real thing that fiber succeeds in doing at the current time, aside from moderately increasing speeds beyond that which can be supported by an all coax system (true, and various other architectural benefits as well), is to boost the expectations of end users. This can be laid to the feet of the MSOs and their ATHM/RR affiliates through their enormous levels of hype that they've generated over the past two years.
It succeeds in driving the ensuing perceived demands to the next plateau, whether there is a genuine need on the parts of the complaining users to attach to such demanding content levels at this time, and whether such target web sites exist to the extent that would justify it, or not.
Consider, users are currently up tight about "only" being allowed 128k in the upstream direction, the direction that merely warrants a few k under normal surfing conditions. Yet, 128 was deemed an out-of-reach goal for the "downstream," never mind the upstream, just a few months before.
Enterprise VPNs, extranets and in house LANs are no different at the distribution level. In the LAN, this means from the closet to the desk, and in this case, no matter how high the backbone speed happens to be, there are times when server backups are taking place, or data bases are being synchronized and replicated, or Virus scans are taking place across all desktops, that relegate the end users' speeds to those of early day asynchronous interface performance levels. I.e., very slow.
If you want more bandwidth, you can always get it. But it ain't free and it's going to have to be properly administered. In the case of HFC, well, I think we've been over this before, but it's my view that the MSOs continued building out a model that was fashioned after some vague 1994 perceptions which are now rapidly becoming anachronisms. In a perverse kind of way, it's the shortsightedness of the early designs which justifies ATHM's behavior in this situation, not their vision. Comments welcome.
Regards, Frank Coluccio |
