Re: [dhcwg] RFC3315 DECLINE definition

[Simon Hobson <[e-mail]>]

"Mudric, Dusan (Dusan)" <[e-mail]> wrote:

> DHCPv6 is then very light weight:
> - address assignment is not a part of the protocol,

How do you come to that idea ? The primary function of DHCP is to provide IP addresses.

> - address selection is not part of the protocol, and

Correct. The job of DHCP is to issue addresses - it can't know in advance what any client might want to use it for.

> - if the client, using its own logic to select from the offered addresses (not part of the protocol), does not select some or any, the protocol should not report that to the server, and

There can be multiple servers, and no client is obliged to accept an offer that doesn't meet it's requirements. For example, a VoIP phone might be configured to only accept an offer which includes options to configure the telephony aspects of the device.
It's not that uncommon to have two servers on a network - one "general IT" one, and one embedded in the PBX. The phones will only take leases from the PBX, and the PBX will ignore clients other than the phones.
The DHCP protocol specification cannot account for all such eventualities.

> - the fact that this kind of protocol can leave a device unreachable does not matter to the protocol that assigns the address.

"Doesn't matter" isn't the right way to describe it. That there are cases where things won't work is inherent in any general purpose protocol required to work in a flexible manner, with a wide range of network topologies and clients, etc.