Ray, concerning why there should be some concerns, you make some good
points, as did Teddy (to whom I was remiss in not replying, after he posted
that excellent framework of grievances which I had requested - Thanks,
Teddy.)
I could add a few more points, which in all likelihood will be counter-
balanced by yet other factors, resulting from yet future and still
unpredictable shifts in the landscape (the tides?) as the 'nets evolve.
Such as the Internet's DNS root servers at some point being made friendlier
to the Europeans and other regionals, thus eliminating the need for every
http hit coming to the states first, simply to target a site a couple of miles
away within the same region of the user. What I've just described, if gone
unfulfilled, will continue to yield enormous amounts of demand to the states.
When fulfilled, however, there will be a dramatic lessening in demand, in
today's bandwidth terms, from that sector. Explan:
When the IETF and other Internet governance bodies (in tandem with the
ITU - oy!) get their collective acts together establishing global peering
points in the UK, Belgium, and elsewhere (and when similar undertakings
are completed in other regions of the world over the next five years) I think
we will see some slackening of the www demand vector in the overall
equation at some point.
This is not to suggest that the capacity will not be consumed in other ways,
but only that some of the economics and the current advantages that are
presently enjoyed due to the legacy of the 'net's architecture might be
momentarily (at least) removed, possibly to some appreciable extent.
Beyond this, the rate at which demand for additional capacity is needed
could be stunted, proportionally.
What the preceding means in straight talk is that when Francois in Paris
wants to do a local lookup of the local wine list which is just down the road
or across town from him, to see what's on special this week, it will get there
through some means which is more direct than his request first having to
come to NY City, and then to MAE-East in Virginia, and then back to the
wine store, and then back to MAE-East again, and ultimately back to
Francois's flat via various untold transiting points.
Now, multiply this scenario by the number and size of transactions that take
place in multinational enterprise VPN activities going forward, and you
begin to get a sense of the potential demand I'm referring to. Since this shift
wont take place overnight, or even fully within the next three or four years,
by any means, the impact wont be sudden and severe, but its eventual and
gradual realization must be taken into account, nonetheless.
Another point I'd like to make - again - is the fact that GBLX is sorely in
need of some IP talent, not to even mention their need for an IP platform
which goes without saying. This is why I think that they must pick up an
ABOV or some other such emerging star in this space. Digital Island also
comes to mind, since I've been reminded of them again recently (yesterday,
in the Last Mile thread), after having ignored them for a while.
If they pick up ABOV, then they might also be looking to pick up MFNX
(the latter, not only for their strand capacity and chutzpah, knocking down
doors everywhere they go), but because MFN has already made a move to
pick up ABOV.
---------
The backbone/last-mile contrast is not clear to me. Actually that is a gross
understatement. It's very very hazy to me at this time, except for the faith I
have that both will be expanding exponentially.All it takes is a national fiber
initiative, either funded by the government or by some private group, to
dramatically upset the balance once again. Whereas today it appears that
there will be an oversupply in the core, the next several years of local
buildouts could easily reverse this trend.
Increases in certain commercial bandwidth consumption trends could wind
up dwarfing the apparent surges in residential consumption, soon, in fact
most casual surfing on the www, when the DWDM-ization of the larger
backbone and VPN SPs matures. Imaging and document archive backups
between data centers, and commercial streaming video applications, to
name just two, could add up to several multiples to what we now have in
the way of today's residential www traffic, for example. And there are
many more than these two applications, obviously, that will be debuting over
the next two years which will demand almost surreal amounts of bandwidth,
from an historical perspective, i.e., compared to what we've been used to
anticipating.
-------
GBLX, in my opinion, needs to upgrade its standing in the ranks of the IP-
capable, by acquiring someone with the requisite skills and platforms
already in place. They don't have the luxury of time on their side to build
their own. Lest they be relegated to the Layer One Chumpion of the Sea
without any means of differentiating themselves at the service level.
I think they'll wind up meeting a part of this need - to get 'net-ified, that is -
if they wind up picking up FRO, but they need to do considerably more than
that, rather soon. All, of course, IMHO. Comments welcome.
Regards, Frank Coluccio |
