Minutes HIP WG at IETF 65

Minutes edited by Gonzalo Camarillo
Based on notes by Pekka Nikander
Meeting chaired by David Ward and Gonzalo Camarillo
Slides presented included in the proceedings

THURSDAY, March 23, 2006, 0900-1130,

Topic: Agenda Bash
Discussions led by: Chairs

The chairs presented the status of the HIP WG items. The architecture
draft will be RFC 4423. The DNS, RVS, and registration drafts have
been ready for publication request for a while, but depend on the base
spec. The mobility and multihoming draft is under WGLC.

The base spec and the ESP draft are waiting for a Beloving-Recorla
analysis. Charlie Kaufman will be performing such an analysis. Another
open issue is the KHI / ORCHID draft. The chairs will talk to the Area
Director to decide how to proceed with it.


Topic: NAT Traversal
Discussion led by: Miika Komu
Relevant documents:
	draft-schmitt-hip-nat-traversal-00.txt

Miika presented the draft. There are two ongoing implementations (NEC
and HIIT). There is a need to figure out the IANA policy on allocating
ports. Can experimental specifications allocate ports?

There were discussions on whether or not it is OK to share the
control/data UDP port number with IKE. It seems that it is probably
OK but it would be good to check.

The IANA policy for the allocation of the 256 reserved SPIs seems to
require the publication of an RFC. The authors will check whether an
experimental RFC would be enough.

The authors will check with the BEHAVE folks whether or not it is OK
to use TURN to implement a server behind a NAT.

More discussions needed on whether an RVS should also relay R1 packets
to facilitate NAT traversal.

There were discussions on whether the LOCATOR parameter can carry
private addresses. Clients may not even know that they have a private
address. More discussions needed.

There were discussions on whether or not implementations should try
first to send control packets without UDP encapsulation. More
discussions needed.


Topic: Rechartering
Discussion led by: Chairs

It was reminded that the HIP WG needs to develop experimental RFCs
that allow to perform experiments with HIP on the
Internet. Research-related issues are outside the scope of the WG.

Initially, two charter items were proposed: how HIP can traverse
legacy NATs and how to have HIP interwork with legacy applications.

There were discussion on whether working on a native API for HIP would
be useful. There was consensus in the group that the WG should work on
the API issue as a charter item.

There was consensus in the room that the WG should add the two
initially proposed items plus the native API work to its charter.

There were discussions about the relationship between HIP and
SHIM6. At this point, both are independent and deal with separate
issues (e.g., locator agility vs. mobility).


The HIP session ended.