Re: [dhcwg] draft-ietf-dhc-dhcpv4-over-ipv6-02 / multiple ipv4 interfaces

[Francis Dupont <[e-mail]>]

 In your previous mail you wrote:

>  The server could have native IPv4 access, used for receiving unicast
>  message to server id.

=> according to my experiments it works only with a physical interface
where renews are received, so it can put a constraint on the topology.
But I tried only one server and not very hard.

>  If we want to address this, we could add a RECOMMANDED setense.

=> RECOMMENDED sentense? Today I am fine with the current wording
even I know it won't be always followed in the real world
(but the non-compliance is with RFC 2131 so please keep the can of
worms closed :-).

>  I agree this is critical. The only case we may break that is if we
>  want to solve the problem in this thread when multiple CRAs with
>  different addresses is unavailable.

=> different addresses in IPv6 is pretty unlikely, even without
multi-interface, so IMHO it is a problem which doesn't need to
be addressed.

>  The reason I insist on the DHCP client for multiple tunnel endpoints
>  with client IDs

=> if endpoint stands for a node then there is one client with one
interface, so no issue.  If it stands for an interface please don't
try to assign addresses to virtual interfaces using DHCP.


>  ... is because we would leave a question mark in the draft
>  if we raise this problem but only mandate the multiple CRA
>  solution (i.e., what if multiple v6 address unavailable)

=> I am not even convinced whether we should have a word about the
problem, so of course I strongly object to put more.

Regards

[e-mail]

PS: I have the draft with some editorial comments in front of me.