April 2005 IEEE 802/IETF Liaison Report
IEEE 802/IETF Relationship History Document
In order to document the history of the IEEE 802/IETF relationship, and describe procedures now in place (such as the New Work list), we have revised the IEEE 802/IETF Relationship History document. Tony Jeffree and Paul Congdon of IEEE 802.1 have said they will circulate it to the IEEE 802 Excomm and Architecture groups for review. It would also be useful to get similar review on the IETF side. Once this is done, the intent is to request that the document be sent to IETF last call, announced on the IEEE 802 mailing list and published as an RFC.
Housley Criteria documentation
At IETF 56 Russ Housley made a legendary presentation describing the security requirements for AAA Key Management.
The security requirements described in that presentation have subsequently been incorporated into a number of IETF documents, including RFC 4017 and the EAP Key Management Framework draft. However, they have not been published as a standalone document. At the ICOS BOF there appeared to be some confusion about the criteria (e.g. which ones are required, recommended, optional). We have also had questions come up in the context of the IEEE 802.16 review request (see below). As a result, it seems like a good idea to publish the criteria as a standalone RFC and a draft is now being prepared.
Dorothy Stanley’s IEEE 802.11 liaison report for March 2005 summarizes the status of ongoing document reviews.
The IETF has requested that IEEE 802.11 review the CAPWAP objectives document, and it is likely that IEEE 802.11 will form an Ad-Hoc committee to prepare a response; target is to approve a resposne at the May 05 Interim in Cairns, Australia.
IEEE 802.21 and IEEE 802.11u have been working on a review of the IAB Link Indications document, as well as the DNA Link Hints document. They did not finish this review at the IEEE 802 March Plenary. Intermediate results are available here.
IETF has received a liaison letter from Roger Marks, chair of IEEE 802.16. The letter requests review of 802.16e by the EAP WG, the MSEC WG, and MIB Doctors. However, the letter does not nominate a liaison from 802.16 to IETF, nor does it provide a timeframe for some of the items. A request for clarification has been sent.
CAPWAP has requested that IEEE 802.11 develop a plan to resolve the authenticator identity issue. The problem is that the 802.11 STA and the AAA server do not have a common view of the identity of the authenticator. An 802.11 STA is only aware of the Access Point BSSID (MAC Address), but an Authenticator can have multiple BSSIDs. While the AAA server receives authenticator identification attributes (NAS-Identifier, NAS-IP-Address, NAS-IPv6-Address), the STA is generally not aware of these attributes, nor is it required that they be securely verified within the 802.11i 4-way handshake. The end result is that the STA cannot determine the scope of a AAA-Key cache and cannot determine when the PMK has been compromised. This has also come up as an issue within the EAP WG. Dorothy Stanley will bring the issue up within 802.11r.
An informal liaison has been established between the IETF BMWG WG and the IEEE 802.11t Task Group. Tom Alexander will be attending both groups. Access to IEEE 802.11t document submissions is available to BMWG WG participants. Access to IEEE 802.11t Recommended Practice Drafts under letter ballot will be provided to BMWG WG participants via the established IEEE 802/IETF agreements, so that BMWG participants can comment on the drafts.
The TRILL Charter was circulated for comment on New Work, and comments were received from participants in IEEE 802, including Mick Seaman (chair of IEEE 802.1 Internetworking) and Donald Eastlake (chair of IEEE 802.11s). Until 802.11s adopts a draft (not expected before March 2006), it will not be clear what the dependencies are. However, it does seem that the existing spanning tree protocol is probably not appropriate for use with 802.11s.
Shortest Path Bridging PAR
During the IEEE 802.1 meeting at the March plenary there was discussion on Shortest Path Bridging. Subsequently a proposed PAR has been circulated, which will be developed during the interim meeting in May in Berlin. It is likely to be on the agenda for approval at the July 2005 Plenary. There was interest from IEEE 802.1 participants, across a range of Bridge vendors. In the closing plenary, there was a vote to pursue the PAR 16-0-2. It provides support for multiple links and it is believed to be backward compatible with existing hardware, and is built on existing VLAN technology. The current direction is to make modest additions to MSTP, not to modify a routing protocol; the general approach could be characterized as “distance vector with handshake” rather than link state. TTL isn’t being discussed. IEEE 802.1 does not believe there are any dependencies on the IETF.
Ballots/Last Calls in Progress
ANSI/TIA standard based on IEEE 802.1AB and RFC 3825
TIA TR-41.4 has invited review and comment on a proposed ANSI/TIA standard tentatively titled “Link Layer Discovery Protocol for Media Endpoint Discovery” (a.k.a LLDP-MED). This new TR-41.4 standard utilizes the IEEE 802.1AB standard as its foundation. The scope of this new development effort is to define an international open set of standard Type Length Value (TLV) extensions for IEEE 802.1AB, as well as necessary management elements, for the purpose of improving multi-vendor interoperability between VoIP devices and LAN equipment. This standard will enable location services, and assist in the management of VoIP networking policies involving Virtual LANs, Priority and Diffserv. The location functionality is based on RFC 3825.
The document is available.
The proposed standard is currently in the TIA ballot phase and will be reviewed at the next TIA TR-41 plenary to be held the week of May 9th 2005. Individuals interested in reviewing the document are requested to submit comments to email@example.com as well as CC to Peter Blatherwick / Editor (firstname.lastname@example.org), no latter then April 29th, 2005.
IETF/IEEE 802 Mailing list
Today a New Work announcement occurs after the IESG has initially considered the charter. This is typically quite late in the process; the IESG will typically vote on the final charter two weeks later. Similarly, IEEE 802 PAR proposals are sent to new work late in the process, often just before the PAR goes out for approval.
It has been proposed that a new mailing list be formed including members of the IESG and the IEEE 802 ExComm and perhaps some other people as well. This list would enable discussion of new work earlier in the process.