On Jun 25, 2007, at 22:28, Christian Huitema wrote:
Suppose that we pre-populate the ip6.arpa tree with synthetic name server records, so that the name server for a given ULA prefix <ula-48>::/48 (ULA-C or not) always resolves to <ula-48>::1 (or any other suitably chosen anycast host identifier). Then DNS look-up will always point to the closest instantiation of that anycast address, or to nothing at all if the prefix is not reachable. Voila, DNS look-up without any central registration...
On Jun 26, 2007, at 02:59, Jeroen Massar wrote:
It would indeed 'solve' the problem given here, but only partially as you still don't have the really important bit: Forward resolving.
I like the idea of asking IANA to define an anycast address for synthetically populating <global-id>.0.0.d.f.ip6.arpa domains with matching NS and AAAA records for the ULA-addressed name servers. Resolvers could easily implement this rule and reduce the load on the public DNS servers. I think the concern about forward resolution and split horizons is orthogonal, and it should be deferred. There is only one horizon required for reverse DNS resolutions of global scope addresses.
I also like the idea of asking IANA to reserve a prefix for RIR's to allocate for non-DFZ use. That effectively addresses the perceived need for a registry to mitigate the risk of birthday paradox collision with ULA.
I'd like to see drafts for both of these (alas, I don't have the time to write them myself). In the process of revising RFC 4193 for the reverse DNS magic, we should also return fc00::/8 to IANA and eliminate the 'L' bit, which I'm on record as saying I think was a bad idea from the start.
-- james woodyatt <[EMAIL PROTECTED]> member of technical staff, communications engineering -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
