83rd IETF ARMD WG session minutes --------------------------------- Location: Maillot meeting room, Le Palais des Congres de Paris, Paris, France Time: 28-Mar-2012, 1510-1610 - Wednesday, Afternoon Session II Chairs: Linda Dunbar, Benson Schliesser Agenda posted at https://datatracker.ietf.org/meeting/83/agenda/armd/ Outline of the Agenda: * Meeting Administrivia * Update on ARMD Milestones (chairs) * BCP for ARP-ND Scaling for Large Data Centers (Linda Dunbar) --------------------------------- Chairs Meeting Administrivia Chairs presented Note Well Thanks to Bert Wijnen for taking notes --------------------------------- Benson Schliesser Update on ARMD Milestones The current deliverables run till Mar 2012. Two items were removed because not enough interest (the 2 Surveys). They are considered done (removed from charter). Problem statement went through last call, but no comments. Chairs assume consensus unless anyone objects. Thomas (author) objects and thinks we need at least 5 people to say YES I read it and support it. AD (Ron) agrees. How many have read the current version? one of the co-authors, and a few hands raised. But no-one on the lists said that they did. So need to ask WG mailing list. Thomas suggests nasty approach to ask people to commit to review. AD will become nastier if he does not see interest. Chair: who will review in 2 weeks? More than 5 committed. ARP/ND statistics draft. Need feed-back So shall we drop it? Unless someone objects strongly we will drop it. Will confirm at mailing list Recommendations draft (milestone Mar 2012) Will be presented GAP Analysis draft (milestone for Mar 2012) Someone MUST step up though. --------------------------------- Linda Dunbar BCP for ARP-ND Scaling for Large Data Centers BCP for ARP/ND Scaling (SEE SLIDES) - SCENARIO 1 - SCENARIO 2 - Static Address Mapping - DNS Based Solution - ARP/ND Proxy Approaches - Overlay Network AD: Is the document BCP or Informational AD thinks Informational is the better choice So say David Black So say Igor Thomas agrees too Maybe we should just say "Current Practices". So do not use BCP acronym (or full) in the doc. WG chair: says charter states we must do a "recommendation". Does that not mean we need to do BCP? AD: NO Thomas: Be practical and don’t get hung up on process WG chair: what does the WG want to do? AD: SHOULD and MUST or should and must WG chair: I am not hearing feedback Linda: Is it more common to have a BCP or just a recommendation? Thomas suggests: Make sure at least 5 people have read it and think it should be a BCP (actually it is not a WG draft yet). WG Chair: So we need to do some more work on this Thomas: put this WG out of its misery Get some people to commit and blame them if they do not deliver We must move quicker to be relevant Anoop: I think this is (still) an interesting problem WG chair agrees. Just that the scope of the WG is limited to “problem” only. There is not much people can do. BUT we must do something or shut down. --------------------------------- Open Discussion Benson: So this was the only item on agenda. Those who raised their hands, please come to me so I can write down your name AD: The Gap Analysis and statistics doc have no author yet or have not shown energy. Shall we just not do them and work on the items we do have activity on? Chair: asks: shall we do that? Thomas: The GROUP does not seem to have the energy SO stop the docs that do not get the energy. Otherwise they may get referenced while they do not really have any standing Chair: So we not only need authors, but we need better content too Thomas: Seems we are not getting that content Chair: Does anyone disagree? Igor: I thought that we last time already agreed to kill this item. Chair: So we're going nowhere with this Thomas: The document as it is is based on simulation and not on real operational workload Igor agrees. As I said I thought we decided to kill it AD: Does anyone disagree with Thomas and Igor? No one disagrees Chair: so we do not do statistics and Gap analysis. So we have the "NOT BCP document" and the Problem statement So 10 have decided to review the Problem statement since very few read the document... AD suggests: AD will sponsor the Problem statement and not-BCP document Then shutdown the WG after this IETF. And all thanks for the good work Thomas: The WG (GROUP) OWES it to the authors to review these 2 documents so they can get published. WG chair: SO pls still come forward and give me your name. ANY last comments D channles xxxx Lots of the work we talked about shows up in NVO3, so that is the place to redirect the energy to. Adjourned.