From: Avi Freedman <freedman@netaxs.com> Subject: Re: CIDR,Sprint and the Big Guys. To: salo@msc.edu (Tim Salo) Date: Fri, 5 Apr 1996 10:18:25 -0500 (EST) Cc: nanog@merit.edu [...]
It would appear that there may also be a strong argument that the tremendous proliferation of [small] ISPs is a significant contributor to the growth of the size of the Internet routing tables.
Perhaps, the [anticipated] consolidation of ISPs will be a significant event in the efforts to control routing table size.
-tjs
Not unless they happen to have contiguous address space.
Oops. I forgot to add "in the long term." Presumably, a smaller number of larger ISPs would be able to justify larger address blocks. But you are correct in observing how difficult it is to go back and clean up earlier allocations.
Seriously, almost every small ISP that I know of gets addresses from its upstream provider(s). Therefore, unless they're dual-homed, they don't contribute to the growth of the size of the tables.
Yes, they may sign up people with IPs in the swamp and route for them, but so may any other ISP/NSP.
I can't imagine that there are more than 300 local and regional ISPs that receive address space from the NIC directly. I could be wrong, of course...
We hear a lot about dual-homed ISPs and ISPs who want to change their upstream providers. It would be very interesting to have some quantitative information about the size of these classes. (Or, are issues related to dual-homed ISPs and changes in upstream providers more theoretical than real?) -tjs