IETF 70 PANA WG Meeting minutes
===============================

Note takers: Victor Fajardo, Jean-Michel Combes


1. WG status and document update

* pana-pana-18 and framework is in RFC editor queue
* pana-fsm will be in WGLC
* pana-ipsec - NEEDs an EDITOR
* pana-preauth - in progress
* pana-mib does not exist and NEEDs an EDITOR

expired docs:
  - cxtp
  - mobopts
  - aaa interworking
  
pana-snmp
  - discontinued
  
  
2. DSL Forum Requirement Analysis

* To support DSL liason request for replacing PPPoE
* DLSF have sent a requirements list of 21 requirements
* All requirements can be fully satisfied by PANA except
  for IP Auth-6 which is kept open due to lack of experties
  in DSL technology to confidently answer the requirement
* RFC4508 already talks about PANA as candidate for DSL deployment
* Analysis is currently in PPT format. We will request that the
  analysis be included in the liason reply to DSL Forum
* Conclusion
  - PANA is specifically designed for problem presented by DSL Forum
  - PANA statisfies DSL Forum subcriber auth requirements
  
Alper   : Questions or comments on the analysis ?
Mark    : I do think you should communicate this conclusion to DSLF
          but put this into an internet drafts instead of PPT 
Mark    : It's the author's conclusion and not a WG conclusion (i.e. concensus)
Alper   : If there is a consensus in the meeting, is there a WG concensus?
Mark    : No, this is just a feeling. Needs of a consensus on the ML.
Alper   : OK, we'll ask on the ML.
Mark    : Slides are may not enough, get an internet draft before
          sending as part of liason response process.
          The normal procedure before sending to the DSL Forum
          is to have an internet draft and get a WG consensus.
Raj     : Since there is a consensus in the PANA WG can we send this
          document to DSL Forum
Mark    : Either email or ID is also good. But in the name of
          expediency, you can try to do it with slides but
          may have problems with legal issues and disclosures.
          Send email to WG based on slides and work from there.
Subir   : Can create this slides plus a quick draft to submit for
          this format
Mark    : Reiterate that for years I wanted a solution. It's not
          about killing pana. Its about the DSLF and vendors proposing
          a DHCP solution and so I threw my weight behind DHCP.
Raj     : We did recognize your work and this is the first time a
          PANA WG meet with base protocol in the editor queue.
          All we want is to propose to DSLF that there is already
          a protocol that meets thier needs.
Subir   : IETF should do the right thing and provide the correct solution.
Alper   : The answer to the question is obvious which is PANA.
Mark    : This is why your going to send a message back to DSL forum
          on why PANA should be used. The outcome of the int-area is
          bigger than the PANA WG.
Bernard : DSLF is asking for request for proposal. Is it to give them
          everything and let them figure it out ?
Mark    : No, the first request was what does IETF have. The second request
          was DSLF has DHCP will you standardize this ? Same happened in
          IEEE.
Bernard : IEEE did decide to send an answer but not sent yet


3. Pre-Authentication

* Changes from 01: Defined 'E' bit instead of 'P' bit
* Updated call flows to be consistent with PANA
* Revised some terms
* Added references
* Issue: Not sure how to deal with MiTM attack because source ip address
         is not protected. Input on this issue is welcomed.
           
Alper   : Other than MiTM attack, is document ready for last call ?
Yoshi   : I think so
Alper   : Please bring this up in the list


4. State Machine

* Updated to synch up with pana-pana-18

Alper   : Need reviewers
Reviewer: Lionel, Julien, Subir ...


5. PaC-EP Master Key

* PaC-EP master key (PEMK) defined in the older rev of the PANA spec
* Suggested to define this in a separate document during LC of base protocol
* Differentiate multiple keys from different EPs
* Security consideration:
   - Channel binding: CB is made at the time of PEMK creation using EPDID
   - KDF can use hoakey 3 party distribution
   
Alper: Need to make sure this document is still aligned with what the
       original base protocol spec intention. Need reviewers, bring it
       back to the list.
       

6. Network Selection

* Also included in the base spec but now separated into another document
* Use a new bit for network selection. 'N'-bit
* Use NAP-avp and flags for NAP selection
* Use ISP-avp and flags for ISP selection

Alper  : Ran out of time, pls take it up on the list


7. Next Steps

Chair:	Re-chartering discussions. 
        Let's identify new items, on the ML and discuss revision of the charter text