Re: "potential new and different architectural approach" to solve the

[Benson Schliesser <bensons () queuefull net>]

On Dec 17, 2010, at 11:23 AM, Joe Greco wrote:

> How effective have variations on hot potato routing been, historically?
> I seem to recall Cogent made lots of noises early on about how they
> could do hot potato routing to encourage peering, but over the years
> that didn't seem to pan out that way.

I can't comment on Cogent...  But, in general: hot-potato reduces network costs but doesn't eliminate them--more 
capacity is still required to carry more traffic.  The goal is to balance out the cost, assuming the traffic is of 
adequate value (or equal value, ideally) to both networks.

Cheers,
-Benson