See now we are back to the catch 22 that is IRR. No one will use it because the data isnt there, and no one will put the data into it because no one uses it. I think the way to get IRR into the real world production realm, is to really drive home the issue w/IPV6. -----Original Message----- From: Richard A Steenbergen [mailto:ras@e-gerbil.net] Sent: Sat 7/13/2002 10:20 PM To: Frank Scalzo Cc: Stephen Stuart; nanog@merit.edu; Paul Schultz Subject: Re: No one behind the wheel at WorldCom On Sat, Jul 13, 2002 at 09:21:16PM -0400, Frank Scalzo wrote:
The underlying problem, is that there are no good widely deployed solutions for controlling what the large backbones inject into the routing table at peering points. A large tier 1 deaggregates towards another bad things happen. It would be nice if there was a supportable way to only allow one isp to advertise appropriate routes to another. The IRR stuff is a neat idea but I dont think many ISPs trust it enough to use it to build ACLs.
If everyone maintained current IRR entries, it would work just fine. The real problem is there are still a lot of networks who's idea of filtering their customers is a prefix-limit or a filter you have to call or email in manually. The only people who actually maintain accurate IRR entries (other than the occational net kook) are those whose transit depends on it. Trying to convert an existing customer base to IRR is a nightmarish task, some of these large established providers will probably NEVER do it unless there is some actual motivation. Supposidly Level 3 requires IRR filtering on their peers, but do they actually try to enforce this? I know they do an excellent job maintaining their own IRR entries, but I'm certain they peer with people who don't have a current db. Probably not, since the vast majority of their current peers don't meet their current peering requirements. :) -- Richard A Steenbergen <ras@e-gerbil.net> http://www.e-gerbil.net/ras PGP Key ID: 0x138EA177 (67 29 D7 BC E8 18 3E DA B2 46 B3 D8 14 36 FE B6)