Re: [dhcwg] I-D Action: draft-ietf-dhc-dhcpv6-client-link-layer-addr-opt-02.txt

["Gaurav Halwasia (ghalwasi)" <[e-mail]>]

Hi Bud,

> If DHCPv6 Server is configured to store or use client link-layer address, it SHOULD look for the client link-layer address option in the RELAY-FORW DHCP message of the DHCPv6 Relay agent closest to the client.

>Although I understand what "closest to the client" means (I use that term myself), is that the correct terminology? It could also be referred to as the inner-most RELAY-FORW message, or the first RELAY-FORW message. Is there even a standard terminology for this? I just glanced through 3315 and don't see anything.

[Gaurav] We could change that to "inner-most RELAY-FORW".

>Also, the server "SHOULD" look for the link-layer address in the closest RELAY-FORW message. What are the ramifications of changing this to MUST? Is there any scenario where a second-hop relay would know the link-layer address, but the first hop wouldn't know it or support this RFC? Do we gain anything by enforcing that a conveyed link-layer address is only valid or to be trusted if it's from the first-hop relay?

[Gaurav] Well the purpose of this spec is to enforce relay agent to add client link-layer address option in the relay forward message. Whereas we don't really want to enforce anything on the server on how it should extract/use this option. 

Thanks,
Gaurav