Re: [BEHAVE] Question on embedded address format in draft-baker-behave-v4v6-framework-02

[Rémi Després <[e-mail]>]

marcelo bagnulo braun  -  le (m/j/a) 4/22/09 9:26 PM:
> Rémi Després escribió:
>> In my understanding, the following constraints result from the fact 
>> that IPv4 mapped addresses have been artificially excluded from valid 
>> destinations in IPv6-only environments:
>> - IPv6-only hosts MUST send ONLY recursive queries (as most host do, 
>> but so far without being obliged to).
> well, they can implement DNS64 in the local resolver, in whcih case they 
> can do non recursive and synthesis locally (that requires host changes 
> indeed)
> 
>> - If these queries are not sent directly to DNS64 servers at 
>> v6only-v4only borders, intermediate servers MUST:
>>
>>   . NOT be in dual stack clouds (there, they should not loose the 
>> ability to respond with As).
> don't understnad this one,could you expand
> 
>>   . NOT handle recursive queries by themselves (although they 
>> typically do it)
> not sure what scenario do you have in mind... areyou talking about a dns 
> server running in a v6 only machine? If so, why it cannot handle 
> recursive lookups? cause the lack of native v4 transport? If so why this 
> is related to DNS64?

1.
I was assuming that the local DNS was not capable to do the translation.
But that may not be worth considering further.

The constraint is then limited to what has already been written, namely 
if I am right:
A. in the IPv6-network-to-IPv4-Internet case, hosts must send recursive 
queries ONLY to DNS64-capable DNS servers.
B. in the IPv6-Internet-to-an-IPv4-network case, "the DNS64 function is 
provided by the end site itself".

2.
BTW, if not done yet, and if I am not mistaken, it's worth noting that 
if DNSs of IPv4-only servers are populated, by hand or by other means, 
with AAAAs containing the agreed WKP and their IPv4 addresses (in 
addition to their As), then these servers become accessible via NAT64s 
*independently of DNS64-capable servers existing anywhere, and with 
IPsec compatibility*.

I don't see which harm a recommendation to do this could do, and the 
benefit can be quite substantial.

With it, agreeing on a WKP becomes even more urgent.

Regards,

RD