[meant to send to all of list, sorry... redirected:]

So is this a good time to ask about the design of a network that concentrates so much of its core service in Auckland? Is TelstraClear's network design any better?

How many regional providers were unable to talk to others in the same region because everything (?) is hauled back to Auckland?

Before you follow this train of thought too far - reverse it. For arguements sake: I'm in Auckland, able to talk to the rest of the world - but nowhere south of Palmy. So what about things hauled back to Wellington? ;-) So the bigger picture question should be 'how many providers have sufficient redundancy to provide services in both locations independent of eachother for any given period?' Obviously theres some outages you just can't protect yourself from - but backup plans and DR plans should incorporate other means of getting info from A to B. Corporates etc with multiple offices can't just be dependant on their PON or DSL. They need to be prepared to make a point-to-point dialup, or an ISP dialup at both ends, in order to be able to communicate critical data at least. (How many companies build their SME LAN's behind a DSL link but business grinds to a halt when they can't get their email because the local exchange had a DSL whoopsie? How hard is it to keep a pay-per-hour dialup plan with an alternate ISP?) Ok, upscale slightly. Corporates and PON with big VPN/WAN environments get more complex. But there are ways. GPRS, Mobile Jetstream? Dialup again, with a VPN client? How many businesses retain 'manual' procedures? Fax machines? Fax-over-Cellular? With ISPs its obviously a different can of worms, but the easiest thing to do is to have what essentially translates into two datacentres - or options on two physically diverse routes to your main datacentre from the other. I don't know if theres enough diversity of choice in the market to let the end usr dictate much of this? The other consideration is that for some businesses - and some scales of bandwidth requirement - its not economical, nor feasible, to have a fully disaster-tolerant network. But its interesting to see how we function when the manure hits the turbine, and to see how many people have actually thought it through that far. Mark.