John Linton
I was offered the opportunity of visiting the 'control centre' of a fair sized ADSL2 network recently as part of an ongoing discussion of, perhaps, sourcing ADSL2 services from another company at some not too distant future time. It was a very 'hush - hush ' visit for reasons known to the people who invited me but of no interest to me. I was also flattered by being invited to look at, what could be regarded as a competitor's operations though I accepted that there was an unspoken 'second agenda' behind the very interesting invitation.
I was very impressed, as I always am, by banks of display screens and network schematics with blinking indicators and performance graphs and other diagnostic screens 'manned' by intense looking people (clearly 'drilled' to make sure they impressed the visitor). I'd seen some of diagnostic reporting tools before as demonstrations when we were looking at buying ADSL2 DSLAMs but it was very interesting to see them used 'live' and to appreciate the benefits of having end to end access to a customer's circuit to find and fix problems.
Very impressive.
I was more interested in the subsequent discussion on capacities and provisioning and their experiences with the difference in usage patterns from their previous experiences with ADSL1 capacities and usage. While having no way of knowing whether I was being told the complete or even some version of the truth it all made sense to me - but as the wise policemen of all popular TV shows would say "all really big and successful lies contain an element of the truth". Not to say that I was being lied to - just to make the point that I wouldn't have been able to tell the difference.
However I can tell a Black Diamond from a Blood Diamond and it did appear to me that at least parts of the back haul were provisioned at 10 gbps and a brief googling when I returned home did suggest that the other boxes were almost certainly 10 gbps Korean copies of the BD 10 gbps NTUs.
So...interesting that at least one ADSL2 provider (albeit at the early stages of their network utilisation) is providing contentionless customer to PoP back haul which some bigger usage networks don't seem to be able to do. I, briefly, discussed their experience in usage by customer profiles and was encouraged that the answers I got were far more in line with what we see at Exetel rather than what other, larger, carriers tell us should be the case.
Given the recent statements in the SPT/TPG merger document about "new ADSL2 connections produce pretty close to gross equals net revenue" perhaps those views are shared by all other ADSL2 network builders (does that mean there's more than one Holy Grail?). If so, then it would explain the desire, of at least one, ADSL2 network owner to attempt to more rapidly build network utilisation via a few wholesale customers who could, each, add 30,000 plus new connections in a short space of time with little or no support and absolutely no additional back end network costs.
While price is, obviously, a major issue in the provision of services it is becoming secondary to the ability of the carrier to actually provision their back hauls at adequate levels. My, not very factual, observations are that the back hauls being provisioned by 'older' carriers are being based on 'historical' ADSL1 provisioning data that simply don't reflect what happens with ADSL2 users - no matter how 'light' the ADSL2 user actually is. Unless there is a major, and rapid, change of view on the required back haul provisioning by at least some of the current carriers my view is that companies like Exetel will not continue to use their services due to the costs of addressing problems that are completely beyond the power/ability of companies like Exetel to address.
It's also very obvious that new infrastructures that start with 10 gbps back hauls and, by definition, are very lightly loaded in their early years are going to be able to provide contentionless ADSL2 far more easily than an 'older' carrier who uses their 'older' infrastructures to carry the newer ADSL2 services.
So maybe there will be some additional, or perhaps at least one additional, options for the provision of contentionless ADSL2 services in the near future?
Though, of course, there's always a catch - at least in my experience to date.
What the catch in this case is - I don't really have any idea other than what happens when the initial 10 gbps circuits are used up and the likely growth in customers above the need for that level has significantly slowed?
Black Diamonds don't cost as much as Blood Diamonds but they certainly aren't cheap.