Minutes IETF103: rtgarea
minutes-103-rtgarea-01

Routing Area Open Meeting (rtgarea)
IETF 103 (Bangkok, Thailand)

===============================================================================
Area Directors:
     Deborah Brungard (db3546@att.com)
     Alvaro Retana (aretana@cisco.com)
     Martin Vigoureux (martin.vigoureux@nokia.com)

Area Secretary:
     Jonathan Hardwick (jonathan.hardwick@metaswitch.com)

Wiki:
     https://trac.tools.ietf.org/area/rtg/trac/wiki/WikiStart

Scribes:
     Jonathan Hardwick (jonathan.hardwick@metaswitch.com)
     Amy Ye (amy.yemin@huawei.com)

Location:
     Chitlada 3, Marriott Marquis Quen's Park, Bangkok, Thailand

Time:
     November 8, 2018, 1610-1810 (4:10pm-6:10pm)

-------------------------------------------------------------------------------

1. Administrivia
   -------------

Alvaro Retana: We have noticed that documents are sometimes coming through
               without having full review by the WG.

Tony Przygienda: It is a good idea to appoint a document shepherd.  They will
                 improve the quality of documents, if chairs lack time.

Alvaro Retana: Yes, but the whole WG is supposed to review the documents, not
               just the shepherds, chairs and ADs.


Alvaro Retana: Please give feedback to the NomCom as IETF appointments are
               being done right now.  E.g. Carlos Martinez & Victor Kuarsingh
               in this room.

Alvaro Retana: We encourage people both to nominate others for IETF leadership
               positions and also to accept the nominations.  This time, 80% of
               the nominees for routing AD turned the nomination down, leaving
               only Alvaro and Deborah standing for re-election unopposed.
               This will be their third term, and it is unknown for a routing
               AD to serve for more than three terms, so we need to bring
               forward more leaders for the future.

Alvaro Retana: The SIDR WG has closed and the SPRING WG has rechartered.

Alvaro Retana: The WG status updates on the RTG Wiki are useful.  In future,
               we will keep a short WG status os each WG up to date on the
               datatracker.  From this, we may be able to auto-generate a
               summary wiki page once per IETF cycle.

Stewart Bryant: (Regarding the AD's suggestion that each WG should define and
                 publish its implementation requirements.)
                One size does not fit all in any WG, so there needs to be a
                chair-exception or AD exception for the corner cases. This is
                so universal that it will apply in every WG.

Martin Vigoureux: I agree with you.  It doesn't hurt to write down explictly
                  the requirements and the fact that exceptions can be allowed
                  on a case-by-case basis.

1.1 Routing Directorate Report
    --------------------------
   Jon Hardwick

Deborah Brungard: Following this IETF, Jon is standing down as the RTGDIR
                  coordinator.  Amy Ye will continue to run the RTGDIR.  We are
                  looking for a volunteer to assist her in doing this.  If you
                  are interested, please let me know.

Stewart Bryant: In your statistics, does 'nits' mean 'ID-nits' or something
                else?

Jon Hardwick: The 'nits' statistic counts any editoral or other minor issue
              which would not block a draft from advancing to publication.

-------------------------------------------------------------------------------

2. DC Routing Update - Status and Technology Updates
   -------------------------------------------------

2.1 Link State Vector Routing (lsvr)
    --------------------------------
   Gunter Van de Velde

No comments or questions.

2.2 Routing in Fat Trees (RIFT)
    ---------------------------
   Jeffrey Zhang

Jeff Tantsura: The way that thus protocol has been developed is a model of IETF
               practice; when we add an extension, we also implement it and
               verify that it works.

Victor Kuarsingh: How do anycast services work on the underlay in the RIFT
                  model?

Tony Przygienda: RIFT is loop free so any path can be taken to anywhere. In IP,
                 anycast is bound by ECMP but in RIFT any path can be taken;
                 traffic can be split too. Also we have fast mobility; Pascal
                 Thubert helped; we changed 6LOWPAN to make a distinction
                 between a 'move' and anycast; normal IP-based protocols cannot
                 do this.

-------------------------------------------------------------------------------

3. Open Discussion / AOB
   ---------------------

Loa Andersson: We got the mail regrading how to update the WG status.  It is
               hard to find the status on the WG page in the datatracker.
               Would it be possible to talk to the datatracker team to add
               that?
Alvaro Retana: Sure, talk to them!
Acee Lindem: I edited the main LSR working group page and put a URL link to the
             status page on the main WG page.  (Under 'additional URLs').
Loa Andersson: I want an additional tab in the datatracker.
Alvaro Retana: OK
Stewart Bryant: The status really belongs at the bottom of the charter - that's
                where you'd expect to find an amplification of the status
                information.
Alvaro Retana: People don't necessarily look at the charters, though. They tend
               to get out of date over time. We are hoping that the status page
               is something that people will keep up to date.
Stewart Bryant: Then just use a pointer at the bottom of the charter.

Jeff Tantsura: The YANG schema mount draft has been submitted for publication.

Stewart Bryant: I think there are too many RTG overlaps in the meeting agenda -
                can we reduce the overlaps?  Does anyone else think this?
Alvaro Retana: Who thinks this?  (About 12 hands up)
Alvaro Retana: WG chairs need to use the conflict lists to avoid this.
Andy Malis: I was _not_ impressed with the Friday experiment. Can the ADs raise
            the concerns to IESG?
Martin Vigoureux: What bothers you, specifically?
Andy Malis: It is a waste of a day. Only 4 small groups meeting tomorrow. It's
            a waste of the resources we are paying for in the hotel. It would
            be better to revert.
Alvaro Retana: There are more than 4 meetings tomorrow, but point taken.
Tony Li: We should reuse Friday. Then we get less meeting overlap.
Greg Mirsky: Agree with Andy & Tony.  Use Friday to avoid conflicts, at least
             in RTG area.
Deborah Brungard: One idea (kicked out) was to have 1 hour session for all WGs.
                  Then people can be more focussed in meetings, and make more
                  use of interims.  RTG area is one of the only areas
                  requesting 2+ hour sessions.
David Black: In transport area, we had a 3 hour meeting and needed 4.
Stewart Bryant: We're in the 21st century, we should use the available data
                from blue sheets etc. to generate some result to figure out the
                possible conflict.
Martin Vigoureux: We need to be careful how to use information such as who
                  attends which WG at what time. There are information privacy
                  concerns.
Alvaro Retana: This time we made sure there were no conflicts between RTGWG
               and other WGs, which will have added to the constraints.

Tony Przygienda: The food in Bangkok was great, but too much!

Jakob Heitz: When do we move on from ASCII RFCs?
Alvaro Retana: The RFC editor has this in progress already.
Stewart Bryant: Colors will be the most useful feature to add.
Alvaro Retana: Take that feedback to the RFC editor.  The IAB manages that
               activity.
Jeff Tantsura: (Speaking as an IAB member) Heather has presented these plans
               twice e.g. colors in documents. Be patient, it is coming.
Alvaro Retana: Please take this discussion to the RFC editor list.

-------------------------------------------------------------------------------

Meeting closed at 1740.
===============================================================================