RE: [dhcwg] Re: WG last call ondraft-ietf-dhc-dhcpv6-opt-dnsconfig-02.txt

On Fri, 2003-02-21 at 11:07, BELOEIL Luc FTRD/DMI/CAE wrote: 
> > Some may scream at this idea, but couldn't we pass an IPv4-mapped 
> > address
> > in there? The DHCPv6 client could recognize this special format
> > to mean this is actually a v4 address?

I think this is a good idea.

> I feel ill at ease with such a solution. How your DHCPv6 client, running
> a node, is aware that there is an IPv4 stack enable on that same node ?
> If it is not, v4-mapped addresses could be harmfull, couldn't they ?

Not really. Why would a network administrator advertise IPv4 DNS servers
on a network with IPv6 only nodes?

Even if there are some IPv6 only nodes on the network, the host resolver
on those nodes will simply try to use the IPv4-mapped address and fail,
and move on the the next one as it would do with any malfunctioning DNS
server.

I don't think the DHCP client has to care whether IPv4 is enabled on the
node (although it could attempt to check that). All it has to do is
configure the DNS addresses and let the host resolver take it from
there.

	MikaL

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg