On Tue, 18 Oct 2005 10:49:36 +0100 Michael.Dillon@btradianz.com wrote:
I reread this and still don't see how geographical ip address allocation is going to work if typical customer connections are network-centric and any large area has number of competitive access providers
Inside the city, you see lots of longer prefixes from that city's netblock. Outside the city you see only the single aggregate prefix.
The only way I see that geographical addressing might have some advantage is if the area is covered by large monopoly that connects everyone else there
Monopoly? Not necessary. Yes, you need to have universal exchange of local traffic in the city but that can happen through private interconnects and multiple exchange points. No need for a monopoly. The major change is that providers which participate in geotopological addressing would have to interconnect with *ALL* other such providers in that city. This would mean more use of public exchange points.
Also, I think it makes sense to have a second regional layer of aggregation where you group neighboring cities that have a lot of traffic with each other. I think this would result in no more than 20-30 regions per continent.
--Michael Dillon
I think that levels of multi-homing are likely to develop for small entities : Multi-homing-0 : You have two or more connnections, but no real sharing of information between them. (I have this now at home, with a Cable modem and dial up for backup. They always appear to the outside as two disjoint networks, and in fact never overlap in time.) I would argue that the vast majority of residences and small offices are are likely to fall into this category; the goal is really internal failover from the preferred provider to the secondary, with automatic renumbering courtesy of DHCP. Multi-homing-1 : You have two or more connections, but can do no traffic engineering, and have to assume an equal preference for each connection. Say there is some sort of geographical or topological prefix. From the outside, they could all be viewed as "belonging" to some preferred carrier, or to a local exchange point, or a protocol could be created to do some sort of topogolocal or geographical provider discovery. It seems to me that this means accepting some sort of hot potato routing, and also some interaction between providers. (The routing would go something like, this is a packet for Clifton, VA; Cox and Verizon cover Clifton Virginia; pick one of these and give it to them and let them worry about the details.) Of course, this scheme it would be highly likely in such a scheme that outbound and inbound traffic for the same flow could use different providers. Multi-homing-2 : What we would now consider as multi-homing, with full control and full BGP. Why would you want Multi-homing-1 ? Well, it should be cheaper than MH-2, with no user administration but you should still get some load balancing and also fast failover if a circuit goes down. That would more than meet the needs of most home offices. If BGP table growth is an issue, I think that some sort of MH-1 is inevitable. I think that inevitably means some sort of geographical or topological based prefix, less-than-optimal routing for at least some packets, and much less user control compared to MH-2. Regional exchanges might be nice, but are not necessary. Regards Marshall Eubanks