Re: Policy Statement on Address Space Allocations
At 11:22 AM 26/1/96, Sean Doran wrote:
| > We just have some differences of philosophy -- you think | > that RIPE really can persuade people into having only | > 1024 announements (preferably far fewer) in 195/8, and | > I don't. That's all. | | OK. I call this a challenge but you won't let me try ;-).
You and Randy Bush seem to be reading each other's minds.
He has proposed this in a way that is very interesting, and which I will think about.
There is a bad failure mode to consider that even a badge afterwards won't make any more attractive.
Mostly it's "what on earth do we do if we cross the threshold of 1024 prefixes in 195/8?" to which I see no easy answer that doesn't involve inflict enormous pain on people with old, established long prefixes in 195/8.
There is at least one very simple response. Set up some deviant CIX, say IX195-8, let everyone with a shortish 195/8 prefix connect to it either directly through their own provider, or indirectly through some tunnel, and have IX195-8 announce reachability of 195/8. That is, in short, altern topology to meet addresses when the converse is too hard. KRE detailed that for the general case, but it would be even simpler in the case of RIPE, since all the allocated network numbers are in the same geographical area. Christian Huitema
To some extent, isn't this how the Amateur Radio folks carve up the 44.*.*.* network? It might be an interesting experiment to use another class A net, sort of like the recent 39.*.*.* (or was it 38?) subnet experiment for such things as web farms, etc. that don't need large allocations, but could really benefit from multi-homing. Ed On Mon, 29 Jan 1996, Christian Huitema wrote:
At 11:22 AM 26/1/96, Sean Doran wrote:
| > We just have some differences of philosophy -- you think | > that RIPE really can persuade people into having only | > 1024 announements (preferably far fewer) in 195/8, and | > I don't. That's all. | | OK. I call this a challenge but you won't let me try ;-).
You and Randy Bush seem to be reading each other's minds.
He has proposed this in a way that is very interesting, and which I will think about.
There is a bad failure mode to consider that even a badge afterwards won't make any more attractive.
Mostly it's "what on earth do we do if we cross the threshold of 1024 prefixes in 195/8?" to which I see no easy answer that doesn't involve inflict enormous pain on people with old, established long prefixes in 195/8.
There is at least one very simple response. Set up some deviant CIX, say IX195-8, let everyone with a shortish 195/8 prefix connect to it either directly through their own provider, or indirectly through some tunnel, and have IX195-8 announce reachability of 195/8. That is, in short, altern topology to meet addresses when the converse is too hard. KRE detailed that for the general case, but it would be even simpler in the case of RIPE, since all the allocated network numbers are in the same geographical area.
Christian Huitema
Ed Morin Northwest Nexus Inc. (206) 455-3505 (voice) Professional Internet Services edm@nwnexus.WA.COM
On Mon, 29 Jan 1996, Ed Morin wrote:
To some extent, isn't this how the Amateur Radio folks carve up the 44.*.*.* network? It might be an interesting experiment to use another class A net, sort of like the recent 39.*.*.* (or was it 38?) subnet experiment for such things as web farms, etc. that don't need large allocations, but could really benefit from multi-homing.
As an amatuer radio operator, I could pick up a class C here in SLC and route it over the internet. I don't think they should carve up another class A just for Web stuff, we are going to than have routing problems again on the net. It would be better that they go to one of thier providers. Yes I know that is not always an option. Christian Nielsen Vyzynz International Inc. cnielsen@vii.com,CN46,KB7HAP Phone 801-568-0999 Fax 801-568-0953 Private Email - Christian@Nielsen.Net PS :)
There is at least one very simple response. Set up some deviant CIX, say IX195-8, let everyone with a shortish 195/8 prefix connect to it either directly through their own provider, or indirectly through some tunnel, and have IX195-8 announce reachability of 195/8. That is, in short, altern topology to meet addresses when the converse is too hard. KRE detailed that for the general case, but it would be even simpler in the case of RIPE, since all the allocated network numbers are in the same geographical area.
I still think it would be worthwhile doing a top-down experiment with this sort of address structure around an easily aggregated geographical area, say the San Francisco Bay Area in northern California. I brought the idea up about 6 months ago and it floundered due to disinterest, but it still seems to be viable. -george william herbert gherbert@crl.com
There is at least one very simple response. Set up some deviant CIX, say IX195-8, let everyone with a shortish 195/8 prefix connect to it either directly through their own provider, or indirectly through some tunnel, and have IX195-8 announce reachability of 195/8. That is, in short, altern topology to meet addresses when the converse is too hard. KRE detailed that for the general case, but it would be even simpler in the case of RIPE, since all the allocated network numbers are in the same geographical area.
I still think it would be worthwhile doing a top-down experiment with this sort of address structure around an easily aggregated geographical area, say the San Francisco Bay Area in northern California. I brought the idea up about 6 months ago and it floundered due to disinterest, but it still seems to be viable.
However, as Andrew w/UUnet pointed out some time ago, you end up providing transit in this way. If the goal is to only announce 195/8, any provider numbered in that block that is dual-homed with this "deviant CIX" and some other provider suddenly starts providing transit for the entire "deviant CIX". I highly doubt that this is desirable. Dave -- Dave Siegel President, RTD Systems & Networking, Inc. (520)623-9663 Network Engineer -- Regional/National NSPs (Cisco) dsiegel@rtd.com User Tracking & Acctg -- "Written by an ISP, http://www.rtd.com/~dsiegel/ for an ISP."
I wrote:
I still think it would be worthwhile doing a top-down experiment with this sort of address structure around an easily aggregated geographical area, say the San Francisco Bay Area in northern California. I brought the idea up about 6 months ago and it floundered due to disinterest, but it still seems to be viable.
Dave replied:
However, as Andrew w/UUnet pointed out some time ago, you end up providing transit in this way. If the goal is to only announce 195/8, any provider numbered in that block that is dual-homed with this "deviant CIX" and some other provider suddenly starts providing transit for the entire "deviant CIX". I highly doubt that this is desirable.
I think that we solved that problem by providing full information for all of the routes to routers within the area, and none outside. If you wish to correct me, feel free. -george
There is at least one very simple response. Set up some deviant CIX, say IX195-8, let everyone with a shortish 195/8 prefix connect to it either directly through their own provider, or indirectly through some tunnel, and have IX195-8 announce reachability of 195/8. That is, in short, altern topology to meet addresses when the converse is too hard. KRE detailed that for the general case, but it would be even simpler in the case of RIPE, since all the allocated network numbers are in the same geographical area.
Christian Huitema
This was proposed and received mild interest about 9 months back. I was able to persuade some people that this was a variant of what was called "provider-based" addressing. It has the interesting trait of also being "geographic-based" as well, for those that can still relate to the old labels. Not enough to take the effort to make it work. I'd be willing to exert some, for any of the existing or planned exchanges, if people thought this would be a viable alternative. --bill
participants (6)
-
bmanning＠ISI.EDU -
Christian Nielsen -
Dave Siegel -
Ed Morin -
George Herbert -
huitema＠pax.inria.fr