From: Brian Renaud <renaud@merit.edu> Here is the breakdown of 198 in the radb, showing the number of route objects registered in 198 for /14 through /24.
prefix length prefix 14 15 16 17 18 19 20 21 22 23 24 ------ ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- 198.* 4 5 39 5 19 30 41 67 98 137 5314
This is interesting, showing how bad assignment has been even _after_ CIDR. 5K sites migrating providers in 2 years. However, I was looking for the older and hoarier problem of 192 and 193. That's where the mostest bang for the buck will occur if we can move long prefix folks into their respective RA blocks. That's also where we will learn about the problems with renumbering relatively older routers and hosts. And that's where folks _deserve_ a break, since they didn't have CIDR allocation in the first place, due to _our_ lack of foresight. So, I'll come by Merit, and see if we can pick a few 192/24s in your AS, and figure out the overall scope of effort that is needed to renumber 192 and 193. Thanks. Merit will be a good starting point, since they already made their sites renumber out of net 35 about 6 years ago! There will be experience which can be built upon. Bill.Simpson@um.cc.umich.edu Key fingerprint = 2E 07 23 03 C5 62 70 D3 59 B1 4F 5E 1D C2 C1 A2
And that's where folks _deserve_ a break, since they didn't have CIDR
Well... why not focus on sites that don't deserve a break? How many /24 from 204 and 205 are announced? Probably not many so I could argue that it would be good to make them renumber. It's not like we're shaking up the whole world. Maybe if the core routers are going to start filtering it should be on a schedule like this: Announce now that by Oct 1, 1996 no individual /24 will be routed. Filter 204/24 and 205/24 on Oct 31, 1995 Filter 202/24 and 203/24 on Nov 30, 1995 etc. until, by Oct 1, 1996, all /24 are filtered. At the same time announce that no new /N prefixes will be accepted by the IRR (and make it so). This will encourage renumbering into larger blocks. Pick N to be a reasonable number based on various registries allocation policies. Regardless of any filtering, the IRR should stop taking any new /24 prefixes so that old 198.* nets (and others) don't creep in before their cutoff date comes up. -mark
On Fri, 29 Sep 1995, Mark Kent wrote:
Maybe if the core routers are going to start filtering it should be on a schedule like this:
Announce now that by Oct 1, 1996 no individual /24 will be routed.
This is a reasonable timetable.
Filter 204/24 and 205/24 on Oct 31, 1995 Filter 202/24 and 203/24 on Nov 30, 1995
This is not! The great bulk of the world, especially management types, gets their technical news with a 4 to 6 month time delay from glossy technical magazines. This delay is due to the time required for writers to reasearch and write as well as the time required for the setup and the presswork. For an example of what I mean, pick up the latest copy of Internet World and read through the news pieces. If this kind of thing *IS* seriously contemplated, you have to push the startup of it out at least 6 months and then start writing some serious press releases to distribute to all the major technical media as well as traditional media. I think the magnitude of the problem deserves some press conferences as well. When the media are bombarded by press releases from Merit, NANOG, Sprint, ANS and so on, they will sit up, take notice, ask questions and report to the world what you want to do and why it is neccessary. Then when the technical guru says to his boss, "We have to renumber all 300 machines in the office by next month end". The boss will understand why and the renumbering will actually get done. Michael Dillon Voice: +1-604-546-8022 Memra Software Inc. Fax: +1-604-542-4130 http://www.memra.com E-mail: michael@memra.com
In message <1621.bsimpson@morningstar.com>, "William Allen Simpson" writes:
From: Brian Renaud <renaud@merit.edu> Here is the breakdown of 198 in the radb, showing the number of route objects registered in 198 for /14 through /24.
prefix length prefix 14 15 16 17 18 19 20 21 22 23 24 ------ ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- 198.* 4 5 39 5 19 30 41 67 98 137 5314
This is interesting, showing how bad assignment has been even _after_ CIDR. 5K sites migrating providers in 2 years.
More likely 5K /24s that were nver aggregated in the first place. What leads you to the conclusion that these were ever aggregated and had to punch a hole because they moved? Curtis
participants (4)
-
Curtis Villamizar -
Mark Kent -
Michael Dillon -
William Allen Simpson