Hello, Sorry for the delayed answer, but I had to verify some things. The answer to the first part of your question is "no" you would not be required to become a sub-TLA yourself. You would receive IP addresses from each of the sub-TLAs you're connecting to. Therefore each host would indeed have 3 IPv6 addresses. For routing, however since you'd get addresses from each sub-TLA, you would just be part of their aggregate routes and would not create any new routing entries in the global routing tables. I recommend reading: ftp://ftp.ripe.net/internet-drafts/draft-ietf-ngtrans-6bone-multi-00.txt Unfortunately it seems to be expired and I don't know what the status is of it becomming an RFC. The author is Francis Dupont who's involved in this working group. Francis, do you know the status of this document? Kind regards, Paula Caslav RIPE NCC Poul-Henning Kamp <phk@critter.freebsd.dk> writes: * [ Moderator note: changed local-ir@ripe.net -> ipv6-wg@ripe.net ] * * * * Could somebody take the time to explain how multi-homing works * in this setup ? I have either not grasped it or not found * the place where it is described. * * Lets take a small ISP who have "upstream" connections to one TLA * and two sub-TLA entities. * * Are they required to become a sub-TLA ? * * if yes: * * What if they don't meet one of the formal requirements for * that ? * * Would failing to become a sub-TLA prevent them from opening * "general" peerings with one or more of these (sub-)TLAs ? * (in other words: will these policies discourage and prevent * multihoming ?) * * if no: * * Which IPv6 numbers will they use ? * * How will route-aggregation work in this case ? * * Will each host have 3 IPv6 numbers listed in DNS ? * * Sorry if these are stupid questions, but I'm probably not the * only one being unenlightened in this area... * * -- * Poul-Henning Kamp FreeBSD coreteam member * phk@FreeBSD.ORG "Real hackers run -current on their laptop." * FreeBSD -- It will take a long time before progress goes too far! * * *