The flexibility of IPv4 routing
According to this article: <http://www.renesys.com/blog/2006/04/tracking_plane_flight_on_inter.shtml> Connexion uses BGP and the Internet routing table to implement IP mobility: | So how did they solve it? They assigned a /24 (256 globally visible IP | addresses) to each plane. They announce that network from the origin | site (in my case, Europe since I took off from Germany). When the | plane is between the two satellites and in view of each, it is | programmed to re-connect to the North American satellite. So traffic | is always getting to the ground the fastest it can, minimizing | latency. Certainly a cool hack, although I'm not sure if the leak of the /24s to the global routing table is intended. It's not necessary to minimize latency. Perhaps Connexion does not operate an intercontinental AS, though -- which would mean that two planes using different head stations could not communicate with each other at the IP layer. Anyway, I wish you guys could bring this kind of flexibility to the IPv6 world. 8-)
On Thu, 2006-05-11 at 20:53 +0200, Florian Weimer wrote: [ .. story about routing .. ]
Anyway, I wish you guys could bring this kind of flexibility to the IPv6 world. 8-)
Too bad a lot of people seem to mix up Address Space with Routing Slots. The RIR's are for the first, not the latter (yet). RIPE and all the other RIR's assign/allocate _Address Space_ to their clients based on their need. How/if that _Address Space_ is actually _Routed_ is not of any concern to the RIR's, that is the task of the one who received the _Address Space_ to make it be like they require it. If you want to make organisations use less routing slots then let them pay for it (business opportunity) and/or filter them out. Unless in the RIPE NCC case RIPE membership demands from RIPE NCC that they start acting as the ITU-TT does. But that is nearly impossible fortunately unless some big ISP's put their hands together and force all you little folks out of the routing tables. (Which will result most likely that small ISP's team up and we get a very cool split internet, but at least with globally unique address space, pfew ;) Greets, Jeroen
* Jeroen Massar:
On Thu, 2006-05-11 at 20:53 +0200, Florian Weimer wrote: [ .. story about routing .. ]
Anyway, I wish you guys could bring this kind of flexibility to the IPv6 world. 8-)
Too bad a lot of people seem to mix up Address Space with Routing Slots.
The RIR's are for the first, not the latter (yet).
For IPv4, there are prefix filtering guidelines, and the RIR assignments are compatible with them. Otherwise, you'd end up with PI space (or even provider allocations) which are not globally reachable to a significant degree. So the two concepts are not entirely separated -- and one stated goal for IPv6 is to mesh them together ("better aggregation").
RIPE and all the other RIR's assign/allocate _Address Space_ to their clients based on their need. How/if that _Address Space_ is actually _Routed_ is not of any concern to the RIR's, that is the task of the one who received the _Address Space_ to make it be like they require it.
I wonder how one would get along with a /23 PI in 62/8. 8->
But that is nearly impossible fortunately unless some big ISP's put their hands together and force all you little folks out of the routing tables.
See, in the IPv6 world as envisioned by you (and others, of course), these little folks are called "waste" -- see <439442BE.2010509@unfix.org>.
participants (2)
-
Florian Weimer -
Jeroen Massar