Re: [mif] Comments on Improved DNS server selection

This part of the discussions reminded me of the very first version of this draft (October 2008)
http://tools.ietf.org/id/draft-savolainen-6man-fqdn-based-if-selection-00.txt
that simply used:
- DHCPv4 Domain Search Option number 119 
- DHCPv6 Domain Search List Option number 24 
to choose between interfaces... 

But then things started growing, as those were not ok to use and as features crept in:)

So is this proposal essentially to go backwards a bit and have option format very similar to Search List Options?

Having only name in the option actually requires having DNS Proxy in figure 2 scenario of draft-ietf-v6ops-multihoming-without-nat66), while having DNS server address in the option makes it possible for hosts to bypass the DNS proxy and talk directly to correct DNS server.

E.g. OPTION_DNS_SERVER_SELECT could have address of this proxy, while new options would point directly to DNS servers with knowledge of private namespaces.

Best regards,

	Teemu

> -----Original Message-----
> From: mif-bounces@ietf.org [mailto:mif-bounces@ietf.org] On Behalf Of
> ext Dave Thaler
> Sent: 28. maaliskuuta 2011 14:22
> To: mif@ietf.org
> Subject: [mif] Comments on Improved DNS server selection
> 
> My thoughts after the end of today's WG discussion...
> 
> I don't yet buy that there's any real world scenario we care about
> where there's one interface that wants to configure different
> DNS servers for different namespaces.
> 
> If you want to do that, you have to deal with hosts that don't
> yet do whatever the new option would do, and to handle them
> you'd hand out the IPs of a DNS server inside that ISP that then
> uses the more detailed information when it does recursive queries
> for example.
> 
> And once you have that, you don't need DNS IPs in the DNS suffix list.
> 
> -Dave
> _______________________________________________
> mif mailing list
> mif@ietf.org
> https://www.ietf.org/mailman/listinfo/mif