Hello Martin,
"LVLT has good management, it is their business model which I question."
I take it you're referring to Level 3's (Crowe's) continued references about horizontalism and their dependency on cols. In the bubble state this was a no-brainer. In more staid times, however, the survivors seem to be vertical players, although one should not jump to the conclusion that a direct relationship necessarily exists here, because other factors come into play having to do with the dominant payer- and monopoly- statuses of many of those vertical players. In any event, this vertical-horizontal thing could wind up being pendulumatic, just as the private line vs public network one is. I suspect that L3 might be looking to leverage their horizontal foothold, wherever it may exist, into more vertically aligned situations going forward. No [p]layer is an island.
[Although some players in the vendor space (Juniper and one or two others come to mind) don't seem to be having a problem in this respect. Not at the present time, in any event.]
"Do any of these players that would acquire these Chp 11 assets run traffic over GX? Will their acquisition of these assets impact GX revenue-model WRT signing up new subscribers of bandwidth?"
Let me turn that around a bit. If a failed entity had previously contracted bandwidth on GX, would their acquirer of assets and subscriber accounts honor those commitments? Here one can see a benefit to up-front IRU payments, I'd imagine, since those are already covered. Or, are they really?
"When these bandwidth deals are arranged, there is apparently a sizeable investment to set up the system to run over a particular carrier. I would expect the costs are sufficient to prevent too much hopping around, as well as contractual arrangements to prevent hopping around from one carrier to another."
The level of drudgery and expense that goes into arranging these bandwidth deals varies from situation to situation - they are case-specific - and in general are in a state of flux, trending to more automated methods as time goes by, this does not apply for all takeers, yet.
At the present time super-T3 deals still require long, sometimes very long, lead times, and require the contracting of "tail section" loops from ILECs and CLECs to the long haul/subsea carrier of choice. For T3s and above the process can take anywhere from 2 to 3 months to a year, depending on local bandwidth and transport node inventories at the local level, and other parameters having to do with the user's proximity to fiber routes.
And there are the CPE investments and engineering costs that must be taken into account at the user level, as well, which I inferred you were also alluding to. The foregoing applies to SONET based provisions - the ones that are not assisted throug optical switching platforms, since some are** - where users must make accommodations over discreet loops. As such, these are still very much a manually intensive set of processes.
[** Begin Sidebar: As an example of one such pop that is 'entirely' supported by optical switching _at_the_carrier_handoff_level_, see the May 2001 Cook Report on Internet executive summary below, where Cook interviews Alexander Muse, CEO of Layer One out of Texas. L1 employs the Ciena Core Director to facilitate the type of colo transfers that I've been alluding to at the granularities that we're discussing:
cookreport.com
It's because of this shift in the model of how carriers need to hand off to one another at the colo and carrier hotel level, in part, that others, such as Corvis and a growing number of startups, are also now incorporating grooming at the T3/STS-1 level, and doing so in a manner that is transparent at the optical level, for all practical purposes. In reality these techniques require the use of o-e-o conversions in backplanes and in switching fabrics. But these are taking place behind the scenes. End Sidebar]]
---
Like I stated above, however, these circumstances are in a constant state of flux. Newer modes of transport and provisioning that employ exchange points, for example, lend themselves to a more 'liquid' form of provisioning, where they exist. For instance, if a large user, or a subscribing carrier, has a large-enough pipe into a carrier hotel or into any of a number of different types of exchange points, then switching to an alternate carrier on the subsea side could be a simple matter of keyboard execution, after some record keeping and fiduciary requirements are fulfilled. Of course, this also assumes that the subsea carrier or one of its transport partners also has a presence in the colo.
Even where they have an access pipe that is only sized to fit (as opposed to being an oversized pipe for future growth purposes, such as a GbE loop extension), if it's going to an exchange point or carrier hotel the user stands a much better chance for an expeditious transfer to a different carrier than if they are served by a discreet loop between themselves and the former carrier of choice. Here we're talking a matter of days or weeks, possibly less than a week in the case of the oversized pipe, compared to months or quarters to do it through traditional methods.
"I think we are mostly dealing with attracting new customers when considering how the failures of some MNOs will impact GX."
This may be an optimistic approach (glass half full?), on the whole, since the net outcome would depend on how many STMs were lost in the first place, by those carriers defaulting, if they were not covered by the upfront payments for IRUs that I referred to above.
In your paragraph that starts with:
"Later, we see partial validation of the perceived..."
... are you referring to the NAP of the Americas? Do you have a link that points to this story, or that enumerates and expands on the circumstances you've stated?
"... a question just came to mind: how can a company easily utilize more than one MNO for their needs? ...can a company, say, readily employ more than one MNO? Technical issues as well as volume-sales issues would be involved..."
Large enterprises and carriers routinely use multiple providers, regardless of the self-healing capabilities that are supported by any one of them. Even GX, itself, through its IXnet subsidiary, participates as a user on other subsea cable consortia facilities, even to places where GX go to. I suspect that the latter is true in part because the communities of interest (financial trading consortia) that IXnet caters to wouldn't have it any other way, which should serve as a salient point, itself.
The ease with which this can be done is facilitated by the use of standards-based transport protocols (which don't care who the provider is) and by colocation methods that I've already alluded to, above. Yes, some economies are lost, and management is more complicated when dealing with multiple providers, but these are common downsides that network managers have contended with from day one. They are the price for reliability and leverage.
FAC |
