RE: Your Discuss on draft-ietf-dhc-subscriber-id

["Wijnen, Bert (Bert)" <bwijnen@lucent.com>]

Ralph, I am not aware I took a DISCUSS for this, and as far as I
can tell from the ID-tracker, all I did is record a COMMENT
to state the question. (Maybe the tracker does confuse you here
becuase it states:
  [public-bwijnen] [IN IESG DISCUSSION *comment*] ....
For a DISCUSS, it states:
  [public-mrw] [IN IESG DISCUSSION *discuss*] ...

Oh well. Thomas had already answered the recorded question and
we found the answer acceptable.

Thanks for the additional info/explanation.

Bert


> -----Original Message-----
> From: Ralph Droms [mailto:rdroms@cisco.com]
> Sent: Tuesday, August 31, 2004 20:22
> To: Wijnen, Bert (Bert)
> Cc: aaa-doctors@ops.ietf.org; margaret@thingmagic.com; 
> narten@us.ibm.com
> Subject: Your Discuss on draft-ietf-dhc-subscriber-id
> 
> 
> Bert - I see that you had a Discuss on draft-ietf-dhc-subscriber-id:
> 
>    An AAA-doctor reviewed this and has a question:
> 
>      This document looks good. One question: There seems to be a
>      growing number of identifiers (client-id etc) for users in
>      the DHCP space. Is there some rule set to determine which of
>      the attributes and in which order are used when, say,
>      determining if the same or different IP address should be
>      handed to the client? For instance, if the client-id has
>      changed but the subscriber ID stays the same, what do you
>      do? Or is this all left to policy?
> 
>    Do we have an answer?
> 
> I also see that your position is now "No-objection".  Just to be clear, the
> subscriber-id sub-option is an identifier for a subscriber that is supplied
> by the DHCP relay agent, based on the logical port to which that subscriber
> is attached, and not an identifier for a specific device (which is what the
> client-id identifies, as supplied by the DHCP client) or for a user (DHCP
> doesn't have a user identifier, per se).  So, while the subscriber-id is yet
> another identifier, it does not overlap the function of the existing
> client-id or MAC address identifiers.  The subscriber-id is not supplied by
> the DHCP client and requires no configuration support in the client.
> 
> The circuit ID and remote ID relay agent sub-options are close in function
> to the subscriber-id sub-option.  The motivation for the subscriber-id is to
> provide late binding between the subscriber and the identifier tag, so
> changes in the subscriber port assignment can be made without reprogramming
> the DHCP server. This function is useful in the case where the subscriber
> connection management is in a different administrative domain from the DHCP
> server (e.g., MSO owns the subscriber connection, ISP runs the DHCP server).
> 
> In any event, policy in the DHCP server governs the use of the various
> values in the DHCP options and relay agent sub-options to determine the IP
> address assignment, accounting and other provisioning.  For example, an ISP
> DHCP server would likely use the subscriber-id sub-option to identify the
> subscriber to which a device belongs and the client-id to identify which of
> the subscriber's devices is initiating the DHCP message exchange.
> 
> - Ralph
>