From owner-routing-wg@ripe.net Tue May 27 11:52:28 1997 From: Tony Barber <tonyb@pipex.net>

Tony, I have checked at random the network below:

...

148.212.1.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.2.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.3.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.4.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.5.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.6.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.7.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.8.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.9.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.10.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.11.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.12.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.13.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.14.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.15.0/24 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.16.0/20 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de 148.212.32.0/20 | AS1292 | ITESM Campus -- Monterr | Universidad Autonoma de

This kind of non-aggregation is what is really hammering the routers on the Internet.

Maybe we at RIPE need to give some serious thought about pan european filtering on address space ?

i.e nothing > /24 Only natural 'old class B's' ??

--Tony

zsako/banknet $ whois -h whois.ra.net 148.212.1 route: 148.212.0.0/16 descr: Universidad Autonoma de descr: Coahuila descr: Jose Cardenas e Hiidalgo s/n descr: Col. Republica Ote. descr: C.P. 2500 descr: Saltillo descr: Coahuila descr: MEXICO origin: AS1292 comm-list: COMM_NSFNET advisory: AS690 1:1800 2:1239(144) 3:1239(218) mnt-by: MAINT-AS1292 changed: nsfnet-admin@merit.edu 950506 source: RADB route: 148.212.0.0/16 descr: UADEC origin: AS1292 advisory: AS690 1:1800 2:1239(144) 3:1239(218) notify: hostmaster@sesqui.net mnt-by: SESQUINET-MAINT-MCI changed: sob@sesqui.net 950426 source: MCI

This kind of non-aggregation is what is really hammering the routers on the Internet.

In this particular instance there is no doubt you are right. The RRs also show a difference between what is announced and what is registered.

Maybe we at RIPE need to give some serious thought about pan european filtering on address space ?

i.e nothing > /24 Only natural 'old class B's' ??

I may agree with the first part (nothing longer than /24), however I feel it would not be wise to filter based on "natural" size of networks. This would prevent us from using the "Class B" address space in the same manner as any other public address space. The same approach caused the problems with 39/8, as far as remember... I think the RIPE NCC may want to allocate sometime in the future from the "Class B" address space, just as we now have allocations from 62/8. Filtering on "natural" size would make this impossible in the case of small registries. A better aproach would be in my view to filter based on RR information (like ANS). I must admit this latter approach is much more difficult to implement and at the same time it does not solve the problem of people also registering crazy routes (not only advertising them). :( Regards, Janos