> I've submitted a new revision of the rfc2462bis draft: > draft-ietf-ipv6-rfc2462bis-00.txt
on the issue of DAD and loopback suppression outlined in appendix A. I've seen a couple of cases in the past where DAD packets could be looped back by the network. this could happen if there is a temporary loop in a bridged network. I've also seen it where parallel links are used between two boxes, e.g Etherchannel. this is something which is very difficult to solve with the current spec. the only alternative is to filter out DAD packets with the same link-layer address as the receiving interface. stopping duplicate MAC address checking from working. one solution could be to add a optional identifier option to DAD NS packets, so that the sender can recognise the case when its own packets are looped around. too big a change for 2462bis? /ot -------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
