Re: [radext] Proposed new charter text

[Peter Deacon <[e-mail]>]

On Fri, 13 Mar 2015, Stefan Winter wrote:

> as you may recall, we discussed new charter text at IETF91, which was
> unanimously accepted in the room.

> We still need to verify consensus on the list, so below is the proposed
> updated charter text. This mail starts a two week comment period on the
> charter text. Please send your comments to the list until 2015-03-27
> 2400 UTC.

> - CoA proxying.  RFC 5176 permits proxying of CoA and Disconnect
> messages, but makes no provisions for how that is done in a roaming
> environment.  This work item will provide descriptions of how to use the
> Operator-Name attribute in a roaming environment to proxy CoA packets.
> It will also define a new attribute which defines an opaque NAS
> identifier which can be used to uniquely identify a visited NAS, and
> whose value will not be modified when proxying, as is done with
> NAS-Identifier and NAS-IP-Address.

Recommend following text:

- CoA proxying.  RFC 5176 permits proxying of CoA and Disconnect messages, 
but makes no provisions for how that is done in a roaming environment. 
This work item will provide descriptions of how to proxy Dynamic 
Authorization messages using realm based proxy.  It will also define a new 
attribute to assist the visited network in processing and forwarding 
Dynamic Authorization requests to appropriate NAS.

While I generally support approach and associated draft 
draft-dekok-radext-coa-proxy-00.txt have some concerns and would prefer to 
keep some prescriptive details out of the charter.

regards,
Peter