Minutes of the 5/11/2015 LMAP WG Virtual Interim Meeting DRAFT Participants * AC = Alissa Cooper * BS = Barbara Stark * GM = Greg Mirsky * JW = Jason Weil * DR = Dan Romascanu * VB = Vaiphav Bajpai * VL = Vic Liu * VU = Vlad Ungureanu * TC = Tim Carey * JL = John Luciani * PE = Phil Eardley * * Agenda - Note Well - WG Status - Information and Data Models - REST Style LMAP Proposal - lmapd Implementation - Next Steps * WG Status - RFC 7536 Large-Scale Broadband Measurement Use Cases - RFC 7398 A Reference Path and Measurement Points for Large-Scale Measurement of Broadband Performance - Framework document in RFC Editor queue - Information Model WG document - Yang Data Model WG document - Protocol document not yet adopted (discussions ongoing) * Information and Data Models VL: definitions of parameters in the IM and DM the same? JS: alligned but not thre same JS: need to look at the cases one by one BS: agreeing w/ JS TC: May have multiple data models - supports methods to map datatypes from data models to IM JS: a number of changes proposed, supported on the list, should we proceed to next iteration? TC: Trevor has some concerns what is missing from the original list JS: send an email with the issues and resolutions - taking an AI * REST Style LMAP Proposal VL presented his slides https://www.ietf.org/proceedings/interim/2015/05/11/lmap/slides/slides-interim-2015-lmap-2-0.pptx JS and VL: both agree that one data model is a goal of the WG JS: looking for feedback on unresolved issues in lmap-yang draft VL: Should we have two separate drafts for Report and Control Protocols? JS: their proposal does as little protocol work as possible and lets RESTCONF WG handle the bulk of that work JS: also call home needed for working behind NATs TC: In BBF spec, they use a single Control Protocol but multiple data collection protocols? DR: Please take a look at the protocol proposals and let us know on the list if anything is missing. DR: I would like to have a discussion and decision in about a months on the protocols. DR: Lets discuss during the rest of May and beginning of June the chairs will try to draw conclusions. JW: We need to have enough people participating in the discussion. PE: I like to state that it is required to work from behind a NAT. JW (Chair hat off) Agree and add IPv6 Firewall case as well * lmapd Implementation VU presented his slides DR: How much additional complexity do you expect for the yet missing parts? VU: The timing objects should be simple to add, for the protocols this may be more work. DR: Did we have a discussion why the resolution is in terms of milliseconds? TC: We typically model time in milliseconds. TC: Milliseconds may be too fine grained but seconds may be too course grained. VL: There is an impact on the network. DR: But we are talking about the scheduling of tasks not about the sending of individual test packets. TC: But we need to go below sub-seconds. The MA has right to refuse resolutions it does not support. DR: I am not sure resource controls are important for the information / data model. JS: The question is whether we leave it to implementations and proprietary data models how to control resources or whether we make this part of a standard (or we do so at a later point in time when there is more experience). DR: Please post to the mailing list any details that should be better explained in the WG documents. * Next Steps - Discuss the protocol aspects and the chairs will draw conclusions beginning on June - JS will post a summary of the proposed changes to the information / data model. - DR proposes to meet for 2.5 hours in Prague, please send time slot requests. - Do we need another interim? VL list, TC list, JS list, ... - AC says IETF 93 might be difficult because of her conflict list. Meeting closed at 20:28 (CET)