Skip to main content

Minutes for ISIS at IETF-94
minutes-94-isis-2

Meeting Minutes IS-IS for IP Internets (isis) WG
Date and time 2015-11-05 04:00
Title Minutes for ISIS at IETF-94
State Active
Other versions plain text
Last updated 2015-11-04

minutes-94-isis-2
IS-IS WG IETF-94
Time Slot (120m): Thursday, November 5th, 2015 1300-1500 JST
Scribe: Acee Lindem (acee@lindem.com) 

 - Intro, Adminastriva, Document Status (See slides) 
     Presenter: Chairs (Christian Hopps, Hannes Gredler)

   Les Ginsberg: RFC6822 Errata will be addressed in a BIS version of 
          the  draft. Some change MAY NOT be backward compatibility. Please
          let Les know if you have an MI (Multi-Instance) implementation. 
   Chris: Depending on implementations and deployments, we may limit the
          changes we take on. 
   Chris: There was a poll for adoption from source/dest routing. 
   Alia Atlas: Source/Dest routing may not be the only solution to
         solve the ingress filtering problem. 
   David Lamparter: V6OPS have recommended the source/dest routing as a 
         solution to solve the ingress filtering problem. 
   Alia: Need to fully understand the problem and all the possible 
         solutions. It is a good idea for the WG to look at source/dest
         routing. 
   Chris: Has this been adopted in the Routing WG? 
   Jeff Tantsura: Source/Dest has been adopted in RTG WG. 
   Alia: Need to look at the enterprise use cases and how this changes
         the context.

 - BIER support via ISIS (See slides) 
     Presenter: Antoni Przygienda
     Document: https://datatracker.ietf.org/doc/draft-ietf-bier-isis-extensions/

 - Advertising L2 Bundle Member Link Attributes in IS-IS (See slides)
     Presenter: Ahmed Bashandy
     Document: http://datatracker.ietf.org/doc/draft-ginsberg-isis-l2bundles/

   Pushpais Sarkar: Document doesn't mention usecases. 
   Ahmed: It is in the beginning. 
   Pushpais: Is there really a reason to push traffic over a particular link in
          a bundle? 
   Ahmed: There is a requirement. 
   Pushpais: A reqirement beyond this probing?
   Ahmed: This is needed to do that. 
   Pushpais: There is another way to probe. Why do we need to advertise
        all this information if it is only for probing? 
   Ahmed: If there is a need to "snake" traffic through layer 2 links. 
   Janos Farkas: I know of use cases. 
   Ahmed: Service providers want this. 
   Chris: Who thinks it should be a WG document? Most who read it think it
      should be a WG document. 

 - RFC4971bis (IS-IS Extensions for Advertising Router Info) (See slides) 
     Presenter: Les Ginsberg
     Document: http://datatracker.ietf.org/doc/draft-ginsberg-isis-rfc4971bis/
     Chris: Anyone think this shouldn't be a WG document? No objections. 

 - IS-IS Minimum Remaining Lifetime (See slides)
   Presenter: Les Ginsberg
   Document: http://datatracker.ietf.org/doc/draft-ginsberg-isis-remaining-lifetime/    

   Chris: Do you cover the case of 0 lifetime case (i.e., an actual purge)? 
   Les: Solution is to use cryptographic authentication along with purge TLV. 
   Tony: What is the resolution?
   Les: Purge TLV is added. 
   Chris: Only set to max age if less than max and "AND non-zero" right?
   Les: Yes - If the draft is wrong, we'll fix it?
   Acee: How does the purge TLV factor into this? 
   Les: There are only certain things that can be in an LSP that is being 
       purged and these are covered in the cryptographic hash to pervent 
       undetected modification. 
   Jon Mitchell: Use "SHOULD" instead of "MAY" for whether situation should
       be logged.
   Chris: Who thinks it should be WG document? Most who read it think 
      it should be a WG document. 

 - IS-IS Extensions for Flow Specification
   Presenter: Jianjie You
   Document: https://datatracker.ietf.org/doc/draft-you-isis-flowspec-extensions/

   Acee: IS-IS is not a PE-CE protocol so this use case is not relevant.
   Julien Meuric: Doesn't really think the IS-IS campus use case is enough to
    justify introducing this feature into IS-IS. 
   Jon Mitchell: Is there really a requirement for this?
   Jianjie: We have customers that have IS-IS and want this. 
   Jon: Can we hear from these people on the list?
   Chris: Seems that there is a need to better discuss use cases on the
    list.

 - IS-IS Autoconf (See slides)
   Document: https://datatracker.ietf.org/doc/draft-liu-isis-auto-conf/
   Presenter: Bing Liu

   Acee: When was the adoption call? I thought I supported this more
    than a month ago. 
   Chris: Will check when the adoption call took place and when it was
    accepted. 
   David Lamparter: Adoption was polled on Oct 22nd. 

 - IS-IS Point-to-Multipoint Operation (See slides)
   Presenter: David Lamparter
   Document: https://datatracker.ietf.org/doc/draft-lamparter-isis-p2mp/

   Chris: The draft has been simplified, we just use P2P hellos, it
    just works, yay IS-IS!
   Chris: This was initiated by HOMENET but it is definitely useful in
    other contexts.
   Alia: In INT Area, there will be a discussion of problems with
    multicast over WiFi. From an individual perspective, I'm happy to 
    see the protocol extended to better operate in this environment.