RE: [dhcwg] I-D ACTION:draft-ietf-dhc-rapid-commit-opt-02.txt

["Kostur, Andre" <[e-mail]>]

Items to note:

Section 3.1:

The first step says "The client broadcasts a DHCPDISCOVER mess on its local
physical."  Physical what?  Interface, I presume?

The second step talks about using the client identifier or chaddr plus
assigned network address as the identifier for the lease.  I believe this
puts it in conflict with RFC 2131 (which talks about using a subnet address,
not necessarily the address of the lease), and RFC 3046 which adds in the
Remote ID.  I believe that this draft should probably reference RFC 3046 on
this topic.

Come to think of it, it looks like this draft re-iterates a bunch of wording
that is already in RFC 2131 (such as:

           When allocating a new address, servers SHOULD check that the     
           offered network address is not already in use; e.g., the 
           server may probe the offered address with an ICMP Echo Request.  
           Servers SHOULD be implemented so that network administrators 
           MAY choose to disable probes of newly allocated addresses. 

).  Does this really need to be restated, or would a reference back to the
"normal" DHCP behaviour as defined in RFC 2131 be sufficient?


Section 4:

This states that a client SHOULD include this option in a discover if it's
prepared to perform the DISCOVER-ACK exchange.  Shouldn't that be a MUST
since there is that if clause at the end?