In a message written on Tue, Jun 14, 2011 at 10:20:07AM +0200, Mikael Abrahamsson wrote:
On the AMSIX peering LAN there is more than 100pps of ND traffic (at least there was when we checked). Since they do not do IPv6 multicast intelligent handling (MLD snooping I guess) certain highend (legacy) router platforms run into trouble because all these packets are punted to RP.
Note that an exchange point LAN is a bit of an odd duck. RA's are supposed to be disabled. There is no DHCP. Rather, the ND behavior is casued by people statically configuring BGP sessions and then a participant leaving. So ND (or even ARP) tries over and over to find the missing participant. The thing to investigate here is if ND rate limiting is implemented correctly by the vendors involved, similar to ARP rate limiting. I'm not sure if there are standards requirements that could be in play as well. I'm not sure this has anything to do with the RA/DHCP issues... -- Leo Bicknell - bicknell@ufp.org - CCIE 3440 PGP keys at http://www.ufp.org/~bicknell/