In einer eMail vom 27.06.2008 20:28:28 Westeuropäische Normalzeit schreibt  
[EMAIL PROTECTED]:

Perhaps  it's a lack of imagination, but I seem to find myself in a maze
of twisty  passages every time I head down this path.  I'd like to
explore  scaling alternatives for the routing system that aren't tied to
loc/id  splits, but I have thus far failed to figure out what that might
possibly  be.



Isn't a look at Google map helpful,e.g.showing a path from New York to San  
Francisco. If you scroll from the start to the destination at reasonably close  
zoom you will realize that, in comparison, the internet with e.g. 210 k DFZ  
-routers is  very very very small. And yet has this scalability problem !??  
IMO, Google ashames IETF.  The loc/id-split gets close to the problem, but  
still fails. It  is one part to get to the egress router, another one  to get 
from there to the egress user, and it is not the network layer's business  to 
care about    session layer's identifying points. There, I  guess, there is no 
problem. The problem is inside the network layer.
 
Heiner



   

Reply via email to