Rémi Després wrote: > james woodyatt - le (m/j/a) 3/11/09 3:02 AM: >> On Mar 10, 2009, at 18:20, Fred Baker wrote: >>> On Mar 10, 2009, at 6:03 PM, james woodyatt wrote: >>>> >>>> Yes, it breaks end-to-end addressability, but not really any worse >>>> than the NAT66 standard that everyone will soon be deploying >>>> everywhere at once, because honestly, who cares that NAT66 lets you >>>> manually provision the DNS with pre-translated AAAA records? >>> >>> Or use Stun and DDNS to do so dynamically? >>> how do we deploy DNS records in IPv4? >> >> >> Yes, and with NAPT44, we use NAT-PMP or UPnP-IGD to find exterior port >> numbers and use those with DDNS to register SRV records for DNS-SD. >> It's a trivial extension to those existing protocols to make them >> support IPv6 for the same purpose with NAPT66. This is what I predict >> we will end up doing in the long run. > > I share this prediction: > - With DNS SRV records, dynamically updated with with global addresses > and locally available ports, many problems become simple. > - In particular, DNS proxies are no longer needed. > - It can also work with port-restricted IPv4 addresses if they are used > as an alternative to NAT44 + UpnP-IGD/NAT-PMP.
most applications do not use SRV records. if they did, they would be less predictable and less reliable than they are now because DNS is so often out of sync with reality. it gets even worse because of multi-faced DNS because the addressing domains and the visibility of those DNS records do not match up, and DNS RRs leak outside of the realms in which they're advertised. advice to everyone working in the NAT space: stop trying to second guess application behavior. Keith _______________________________________________ nat66 mailing list [email protected] https://www.ietf.org/mailman/listinfo/nat66
