-----Ursprüngliche Mitteilung----- 
Von: Paul Jakma <p...@jakma.org>
An: Robert Raszuk <rob...@raszuk.net>
Cc: rrg@irtf.org
Verschickt: Mo., 26. Apr. 2010, 13:27
Thema: Re: [rrg] Next pass


On Sun, 25 Apr 2010, Robert Raszuk wrote: 
 
> Your destination (content you are looking for) is more then likely > to be 
> very distributed - even globally. 
 
Perhaps obvious, but note that we can easily have a system that allows for both 
kinds of routing. 


yes,for sure: The amount of extra-churn for TARA info is neglectable compared 
with the permanent increase of the FIBs and RIBs. The size of the required 
TARA-routing table is pretty static.
 
Prefix a -> Routed with reference to some external, universally understood 
space 
 
Prefix b -> Routed without reference to any such space 
 
If it is the case that geographic routing could "factor out" a sizable 
proportion of routes, then that's surely worth investigating? To allow the fact 
that such a system can not be used to aggregate /all/ routes to stymie its 
consideration completely is perhaps short-sighted. 


Note, Geographical data  is used to compose skimmed topologies (zooms) 
properly, i.e. not simple to select the next hop towards the destination's 
geo-location. So any detour is still enabled.
 
More generally, if we have a number of different routing problems to solve, and 
if we are having trouble finding The One Way that solves them all, perhaps then 
we need to consider building a system composed of multiple different ways. 


Topology-aware routing provides a huge multitude more of routes than a DV-based 
system (where you don't know all those detours via  nodes to which you have 
forwarded the BGP UPDATE)
 
Wrt geographic routing: The ITU proposals to delegate to and assign IPv6 
address space via per-country, statutory bodies could be a vehicle for such. 
Course, there is some opposition in places to the very idea of allowing 
delegation via ITU.. 


There is a major difference with E.164 country codes: There you will always 
have ( at the country level) what I call the Istanbul effect.


Heiner
 
regards, 
-- Paul Jakma  p...@jakma.org  Key ID: 64A2FF6A 
Fortune: 
The Fifth Rule: 
   You have taken yourself too seriously. 
_______________________________________________ 
rrg mailing list 
rrg@irtf.org 
http://www.irtf.org/mailman/listinfo/rrg 

 
_______________________________________________
rrg mailing list
rrg@irtf.org
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to