Re: [mif] Comments on draft-mouton-mif-dhcpv6-drlo-00

[maximilien mouton <[e-mail]>]

Hi Simon and Ted,

Le 13/09/2011 20:26, Simon Perreault a écrit :
> Ted Lemon wrote, on 09/13/2011 02:22 PM:
>> Could you please clearly articulate a use case where this option is
>> necessary, and the existing proposed option is not adequate?
>> Please do not mention operating system data structures.   Please do
>> not say "I find it to be the case that.."   Without a use case, I
>> just don't see the point in discussing this further.
>
> +1

First, I think we can say that there is a consensus on the fact that
DHCP should have an option including default router.

Second, why a specific option for default route rather than specify it
in route option?

You can notice that router option and static route option (the stateful
one) are defined separately since rfc 2132 (DHCP Options).
Moreover, rfc 3442 (Route Option for DHCP version 4) has a similar point
of view than what we propose with DRLO.

But beside this history point here is a scenario explaining the need of
specific option for default route:
Imagine a realistic network in which in the same prefix is, a limited
resource device not able to carry more than a default route (see
previous discussion with Alex), a classic device like a laptop with one
interface and a MIF device which want to run route option.
A reasonable configuration is DRLO for 2 the two first type of device
and DRLO+route option for the third (less frequent case).

In the opposit specify specific routes without default route may be
useful for instance for security or accounting reasons.

Maximilien


>
> Simon