[lisp] LISP minutes IETF 93

[Luigi Iannone <[e-mail]>]

Hi All,

hereafter you can find the minutes of our last meeting.
Available also at: https://www.ietf.org/proceedings/93/minutes/minutes-93-lisp <https://www.ietf.org/proceedings/93/minutes/minutes-93-lisp>

Please let us know if you have any comment on them.

ciao

Luigi


---------------------------------------------------------------------------------------------------------------
- Administration        
   Halpern/Iannone
   - Blue Sheets
   - Agenda Bashing
   - Status reports for WG drafts

[Darrel Lewis] will  send in a week the new updated DDT draft.

[Terry Manderson] proposes to re-arrange the agenda to start with WG documents. The WG accepts. 


o WG Documents Update

- LISP LCAF
    https://tools.ietf.org/html/draft-ietf-lisp-lcaf-10
    D. Farinacci

[Joel Halpern] notices that the document does not explain how to use the different types.
[Darrel Lewis] the document should be rewritten to be more readable. 
[Darrel Lewis/Joel Halpern] propose to only put in the document how to make LCAF and a few examples instead of a full list.
   [Dino Farinacci] answers that most of types proposed in the document are described in other drafts (some being expired).
[Joel Halpern] reminds that this draft blocks intro document as LCAF is a normative reference for it.


- LISP Crypto
       https://tools.ietf.org/html/draft-ietf-lisp-crypto-01
       D. Farinacci

[Darrel Lewis] likes the securization of the control-plane but questions why proposing a new mechanism for the data-plane while ESP is already widely deployed and optimised.
  [Dino Farinacci] says that for him the important part of LISP is the control-plane, how the data-plane is made is a detail.
  [Dino Farinacci] in ESP there is some functions that are not necessary and using IPSEC means that it would be more headers and then having MTU issues. 
[? from Cisco] asks how to migrate to a non-security-enabled  to a security-enabled LISP site.
  [Dino Farinacci] this is straightforward: if the LCAF is not understood, you skip it.



o Rechartering Discussion

- Cloud SOHO Services
           D. Saucez

[Damien Saucez] The presentation is a demo of multihoming in SOHO networks with LISP. 
[Dino Farinacci/Albert Cabellos] By experience in public NATed environments and Amazone, it works pretty fine. In more constraints environments (e.g., corporate networks) it is harder as firewalls are usually not open.
   [Damien Saucez] if Skype can pass, then LISP should pass as well, we have to find viable solutions for such environments
[Sharon Barkai] says that what matters is the construction of the overly to go to the Cloud.
   [Damien Saucez] of course, this is the mapping system that make all here. The data-plane is a detail and it may not be necessary to have the LISP data-plane
[Darrel Lewis] this demo fits very well in the multi-homing use case part of the original charter and it should be in the new charter.


- LISP WG Mailing List Summary
    A. Cabellos

- Open mic discussion

[Darrel Lewis/Joel Halpern]  important to separate technical changes and use cases. We have to clearly define what are the changes to put RFC 6830bis and those not to put in RFC6830bis.

[Darrel Lewis]: do not tight the different protocol details and the core (e.g., mapping system, NAT traversal…) to a particular use case.

[Terry Manderson]: proposes to remove the IAB requirement about scalability of routing table problem to concentrate on the core protocol and think about the critical use cases that the WG want to work on.

  [Joel Halpern/Dino Farinacci/Albert Cabellos] agree with Terry

[Ed Lopez] says problem can be separated in two: the case of LISP as end-system, and the one in intermediate system.

[Darrel Lewis] make the distinction between immediate use cases and exploratory use cases.

[Michael Menth] says it is important to have deployment to understand what could be the use case.

[Sharon Barkai] says that what is important is the notion of overlay and that the group should propose a clear interface for “map assisted overlays”. Opening the schema to different solutions and underlays.

[Larry Kreeger]: proposes to get more with NVO3.

[Joel Halpern]: asks attendees to refine categories and send to the list


o Non WG Documents

- LISP Hybrid Access
          https://tools.ietf.org/html/draft-menth-lisp-ha-00
           A. Stockmayer

[Darrel Lewis] Maybe it would be possible to put state in the mapping system instead of in the boxes so to be compatible
   [Andreas Stockmayer] then it is not possible to have the timestamp.
[?] asks if LISP is necessary to do it.
  [Luigi Iannone] the advantage of LISP is that it provides a fully functional system (control-plane + data-plane).

- LISP Multi Tuple EIDs
          https://tools.ietf.org/html/draft-rodrigueznatal-lisp-multi-tuple-eids-00
           A. Cabellos