On Wed, Mar 31, 2010 at 11:26:53PM -0700, Christopher Morrow wrote:
> On Wed, Mar 31, 2010 at 1:55 PM, Dan Wing <dw...@cisco.com> wrote:
>
> > But Remi's point is that those same systems (running Windows XP
> > and IE6) using 6rd will be denied the ability to access content
> > via IPv6. Which removes an incentive for ISPs to add 6rd (and
> > offload the NAT44 they may soon have to install).
>
> I'm not sure this is true:
> o the end-station sends a dns query to the ISP recursive resolver
> o the recursive resolver uses whichever protocol is 'best' for the
> lookup (assuming something like BIND's NS RTT caching happens in the
> majority of cases)
> o if the query goes over ipv6, is for a AAAA, a AAAA response could be
> returned
> o if the query goes over ipv4, is for a AAAA, no response is returned
> and presumably a follow-up A query happens.
>
> Igor/Yahoo are proposing that recursive resolvers which have v6
> transport use it and be rewarded for doing so... if they have a
> longer/worse path over ipv6 there's a good chance the user experience
> will also suffer, so the users should use v4.
but the ISP recursive resolver has no clue what the end-station
(stub) is going to do with the result of the lookup. nor does
if have any idea what proxies or forwarders are in the interviening
path between it and the stub.
it might look at the transport it received the query on (from the stub),
cache that information, and then recursively ask the authoritative
servers
for the data (using whichever transport is "best" for the specific
lookup
to the authoritative server in question), then send the response back
to the stub - weighting the response data based on the transport the
stub asked the query on.
regardless of how I (the stub) ask the recursive my question, (via IPv4,
IPv6, DECNETPhaseV, NCP, bisync...) the recursive should give me what
i ask for ... if i want AAAA answers then give me AAAA answers, even if
you get the query over LU6.2. Don't confuse DNS data with transport.
--bill
>
> -chris
>
> -chris
> _______________________________________________
> DNSOP mailing list
> DNSOP@ietf.org
> https://www.ietf.org/mailman/listinfo/dnsop
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop
