Network expansion based on customer demand (was: Re: [e2e] Fundamental Questions about Router Queue in High Speed IP Networks)

[Anders Bergsten]

David, 

> The false assumption pervading this whole discussion is that it presumes a
> constant network capacity, rather than assuming that the goal of a network
> is to expand to meet all the demand that customers are willing to pay for,
> and that customers will not pay very much for long latency and high jitter.
A follow-up question on that...

I agree that the goal of a network is to expand to meet all the demand
that customers are willing to pay for. Of course, it has to be under the
assumption that the expansion provides revenu. This expansion is the key
to achieve QoS (demand<=supply - whatever technology we use).

But, would you say that we (as an ISP for example) have enough
information to judge where to expand the network to meet the demand that
the customers are willing to pay for? 

As I see it, we only have information about the resource utilisation on
single links to judge the demand. We also have the information about our
customer contracts that gives some hint on what they want - but
generally these contracts does not state anything but the aggregated
bandwidth that the user wish to send. 

What I am especially thinking about is TCP's congestion control, which
might actually hide some information from us - the true demand of a user
might be quite different than what the transport mechanism allows, and
thus what we see in the network as resource utilisation. 

Anders