Minutes of the IRTF HIP research group meeting, March 23 2009, San Francisco

Minutes compiled by Tobias Heer and edited by Tom Henderson

The meeting was co-chaired by Andrei Gurtov and Tom Henderson.  33 people
signed the attendance sheets.

1) HIP Experiment Report update, and discussion on next steps
 - http://www.ietf.org/internet-drafts/draft-irtf-hip-experiment-05.txt

General overview presented by Tom Henderson
  - Overview of the RG-related drafts
  - Overview of HIP implementations
  - Working group updates
  - Future directions
  - Discuss experiments with HIP on larger scale?
  - Implementation and deployment experience should be included in 
Experiment Report

2) Gyu Myoung Lee, update on HIP for object to object communications
  - http://www.ietf.org/internet-drafts/draft-lee-hip-object-02.txt

  (See slides for presentation summary)

  - Robert Moskowitz: This can be done with HIP without modifying HIP by 
seeing objects as services.
  - Pascal Urien: The work looks promising. But a big issue for deployment 
is privacy. Users can be traced everywhere with their HIT. Objects 
aggravate this problem.
  - Tom Henderson: More discussions on the list please.

3) Tobias Heer, HIP middleboxes draft update
  - http://www.ietf.org/internet-drafts/draft-heer-hip-middle-auth-02.txt

  (See slides)
  - Updates od midauth draft
    - Tobias Heer: Comments on the draft?  
    - Tom Henderson: Next steps for the draft?
    - Tobias: We want to finish the prototype for cersion 02 and would 
like to go back to the security discussion then. We certainly need input 
from the community then. 

4) Tobias Heer, HIP services draft
  - http://www.ietf.org/internet-drafts/draft-heer-hip-service-00.txt

  - New draft: hip-serv-id slides
    - Tobias: Input from the community for service types?
    - Andrei Gurtov: Can  object ids be put into the service id?
	- Andrei: There might be some intersection in these works.
    - Tom: Is this part of a general architecture for middlebox-to-host, or 
a more specific negotiation limited to piggybacking on existing exchanges?
Can this be extended to signal the requirement of a certificate / certain 
certificate type?
    - Tobias: Yes. Announcing the need for a certificates was one of 
the original motivations for this work.

5) Oleg Ponomarev, HIT-to-IP mapping service
  - http://tools.ietf.org/html/draft-ponomarev-hip-hit2ip-03

   (See slides)
   - Robert: There should be the address of the RVS in the DNS, not the 
address of the device.
   - Tim Shepard: Are all applications going through HIP?
   - Oleg Ponomarev: Destination HIT determines need for HIP. 
   - Tim: Firefox is not a good example for a application that needs HIP.
   - Oleg: HIP may well be useful for long-lasting connections.
   - Tim: Legacy apps should not use HIP by default. Firefox will probably 
not be a legacy app for long.
   - Oleg: I would like to see HIP in practice, not just in a lab.
   - Tom: Will this be discussed in another meeting here in San Fransisco?
   - Oleg: Just breiefly with DNS ops people.
   - Tom: We should continue discussions this on the list.

6) Andrei Gurtov, hip-sava draft
  - http://www.ietf.org/internet-drafts/draft-kuptsov-sava-hip-01.txt

SAVAH Source address validation with Host Identity Protocol 
   (See slides)
   - Robert: What happens in case of a mobility event?
   - Andrei: There should be a UPDATE between first-hop router and the end-host.
   - Tom:  Will there be discussion at SAVA WG?
   - Andrei:  Yes, briefly.

7) Samu Varjonen, hip-cert update 
  - http://www.ietf.org/internet-drafts/draft-ietf-hip-cert-00.txt 

  - Update on HIP-CERT
     - Samu Varjonen: Any comments on the draft?
     - Tom: We also did some implementation work with X.509 and size is an 
issue.
     - Gonzalo Camarillo: We need to do something about fragmentation. This 
issue keeps showing up all the time. There must be a solution in the RG or WG. 

8) Samu Varjonen, hip-srt (strong password auth of users)
  - http://www.ietf.org/internet-drafts/draft-varjonen-hip-srp-00.txt
  - New draft: HIP-SRP 
    - Robert: Patents seem to be a issue for SRP. There are overlapping patents.
    - Samu: Is there any interst in this?
    - Tom: I think there is interest in associating HIP associations with users.
    - Andrei: Could you use it for objects (reference to talk 1)?
    - Samu: Possibly. 

9) Xiaohu Xu, RANGI proxy
  - http://www.ietf.org/internet-drafts/draft-xu-rangi-proxy-00.txt

  (See slides)

 - Oleg: Is there any difference to LISP?
 - Xiaohu: We borrowed some ideas from it.
 - Oleg: You need massive caching in the name resolver system. What would 
be the size of such cache?
 - Xiaohu: This is a general issue for map-and-encaps.
 - Oleg: To my experience, even one host with GB connection can easily 
overload such system.
 - Tom: This seems to be HIP with hierarchical HITs?
 - Xiaohu: Yes.
 - Tom: So this is a tunelling mechanism for dealing with legacy hosts?
 - Xiaohu: Yes. 
 - Tom: This seems to me to be more than just LISP or map-encap because you 
seem to be translating IDs. Does the Proxy translate the ID?
 - Xiaohu: That depends on your use case. There is no checksum recomputation, 
no translation. 
 - Tim: Is there mobility support in RANGI (just like in HIP)?
 - Xiaohu: Since it is a ID/Loc split approach it can support mobility.

9) Robert Moskowitz, HIP and SIP over IPv6 status
 (see slides)

The meeting concluded after Robert's talk, without questions, due to
the end of the meeting slot.