Minutes for LISP at IETF-96
minutes-96-lisp-1

Meeting Minutes Locator/ID Separation Protocol (lisp) WG
Title Minutes for LISP at IETF-96
State Active
Other versions plain text
Last updated 2016-07-31

Meeting Minutes
minutes-96-lisp

   CHAIR(s):  Joel Halpern ( jmh AT joelhalpern.com )
           Luigi Iannone ( ggx AT gigix.net )

SECRETARY: Wassim Haddad ( wassim.haddad AT ericsson.com )
           Damien Saucez ( damien.saucez AT gmail.com )

        
AGENDA

Session 1/1 (120 Minutes)
=-=-=-=-=-=-=-=-=-

Thursday, July 21, 2016
1620-1820, Afternoon Session III, 120 Minutes
Room Charlottenburg I

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

_ Chairs remind the new charter has started

_ Chairs ask people to comment adoption of draft-boucaidair-lisp-type-iana-01
  on the mailing list.

_ Damien Saucez will step down from secretary at the end of the year,
  volunteers can contact the chairs.


o WG Items

- LISP Yang Model - draft-ietf-lisp-yang-02
	V. Ermagan

	- Vina Ermagan asking the WG to change the title of the draft to 
	cover better the content of the document
  	- Chairs ask to send the request to the list
	- Chairs ask if the Benoit Claise remark has been addressed
  	- Vina Ermagan says new version fixes it, thanks to new online 
	checking tool.



o Non WG Items

- LISP Data Plane Discussion (6830bis)
	A. Cabellos

	- Albert Cabellos remind the rational of "bis" documents.

	- Dino Farinacci says moving mobility consideration to the mobility 
	document would impair its usage as it is not a WG doc.
 	- Luigi Iannone says it can be moved out as mobility has to be supported 
	by the group so the text will go in the document adopted by the group.
	- Luigi Iannone says for him the green parts should just be removed from 
	the document.
	- Luigi Iannone says for him RLOC probing/Clock Sweep should be in Control
	Plane.
	- Dino Farinacci says that if people want to use LISP control plane,
	seeing RLOC probing in control plane document may look awkward for
	people not willing to use the lisp data encapsulation.
	- Luigi Iannone agrees with this point.
	- Richard Li says that the mapping could be used by other groups and 
	propose to discuss it separately.
	- Padma Pillay-Esnault mapping is universal to everything and beyond 
	this WG.
	- Vina Ermagan says the protocol is part of the WG, not the algorithms.
	- Joel Halpern says we need to define the protocol as we need it.
	- Luigi Iannone asks the group to think about the necessity of a IANA
	registry in this context and check draft-boucaidair-lisp-type-iana-01.
	- Dino Farinacci says we have to make more marketing about DDT.
	- Albert Cabellos says that the 6833bis should be only interface
	specification.
	- Luigi Iannone says if we want to show this is generic control plane, 
	what about putting an example with LISP and an example with another encap.
	- Richard Li asks how to talk about the heterogeneous networks (LISP and
 	non-LISP).
	- Joel Halpern says the existing documents already discuss the LISP to
    	non-LISP and vice versa. 
	- Joel Halpern prefers to avoid adding potential future use cases in this
    	particular document.
	- Dino Farinacci says there are interoperability and deployment documents
    	already.
	- Albert Cabellos express his will to be strictly editorial, not adding 
	new things.
	- Dino Farinacci asks implementer to carefully check to see if things 
	have to be changed because of bugs in 6830 and 6833. 


- LISP Header Compression - draft-saldana-lisp-compress-mux-00
	10 Minutes	(Cumulative Time: 55 Minutes)
	J. Saldana	

	- Luigi Iannone asks if for the first approach they tried used port number.
  	- Jose Saldana says it is worth a try.
	- Fabio Maino proposes to look at VxLAN-GPE.
  	- Fabio Maino says it would be worth exploring involving the control plane.
	- Xiaohu Xu asks if side effects on latency has been studied.
	- Jose Saldana answers that it has been studied and that it is better not 
	to multiplex packets from a same conversation.
	- Xiaohu Xu mentions the risk of reordering.
	- Dino Farinacci says it may not happen as RLOC selection is made on the
    	5-tuple.
	- Jeff Tantsura says the a problem is the jitter.
	- Jeff Tantsura comments that it is very different what you can do for 
	a 2Mbps link and a 10Gbps.



- jLISP: An Extensible Java Based Open Source LISP Implementation
	A. Stockmayer

	- Joel Halpern reminds that IETF likes implementation and encourage 
	people to implement.
	- Dino Farinacci asks if the API could be used with JavaScript.
 	- Andreas Stockmayer replies that some work is needed. 
	- Albert Cabellos asks if it is interoperable.
	- Andreas Stockmayer answers it is tested with OOR xTR and OpenLISP 
	control plane.
	- Luigi Iannone asks for a demo.
	- Fabio Maino asks if it is compatible in iOS.
	- Andreas Stockmayer answers it would need the VPN API if not rooted 
	device.



- ComCast IPv6 SR 
	J. Brzozowski

	- Robert Radzuk asks where is LISP.
	- Luigi Iannone says it is for documentation as LISP is trying to solve
	the same kind of problems.
	- Dino Farinacci asks if there are traffic engineering requirements.
	- John Brzozowski says that for the moment no but that it may come in 
	some use cases


!!! NOTE WELL: Order changed compared to initial  agenda !!!

- LISP EID Distinguished Name Encoding - draft-farinacci-lisp-name-encoding-00
	D. Farinacci

- LISP EID Geo-Coordinates Use-Cases - draft-farinacci-lisp-geo-00
	D. Farinacci

- LISP EID Anonymity - draft-farinacci-lisp-eid-anonymity-00
	D. Farinacci

	- Luigi Iannone asks about address the risk of collision.
	- Dino Farinacci did not check yet.
 	- Luigi Iannone says there are already specs to randomly change IPv6
    	addresses.



- LISP EID Mobility - draft-portoles-lisp-eid-mobility-00
		     - draft-meyer-lisp-mn-15
	D. Farinacci

	- Xiaohu Xu asks how to trust ITR.
	- Dino Farinacci uses LISP-sec and registration with keys and you trust 
	the Map-Server.
	- Wolfgang Riedel what's happening when you use captive portals where you 
	are connected but have nothing until you click I agree.
	- Albert Cabellos asks how to register the new mapping.
	- Joel Halpern: no time, talk to the list.



- LISP Predictive RLOCs - draft-farinacci-lisp-predictive-rlocs-00
	D. Farinacci

	- Joel Halpern asks if the assumption is to know the direction
  	- Dino Farinacci: yes
	- Luigi Iannone for any question, go to the list.


- End of the session