IETF 80 ROLL WG Session Thursday, 31st March (1520-1720) Agenda and Slides: http://tools.ietf.org/wg/roll/agenda 1) WG Status (Chairs) WG Update Slides - Excellent progress since IETF79 and the interim WG session. - One new RFC (RFC6206) since IETF 79. - Two documents in the RFC editor queue (draft-ietf-roll-rp and draft-ietf-roll-routing-metric). - Core specification of RPL is complete. - Plan to reserve some time at the end of the session a recharter discussion. - Chairs insisted on the need for the WG to close on P2P work, and start working on the applicability statements. 2) Recommendations for Efficient Implementation of RPL (Phil Levis) draft-gnawali-roll-rpl-recommendations-01 - Phil Levis provided an update on the recommendations draft based on experience. It is also possible that another WG may take on additional implementation items. - Co-chairs highlighted that the objective is to tune the protocol based on recommendations. - Thomas mentioned it may be beneficial to retain neighbor state in order to be more efficient. Phil concurred it would be interesting to have more opinion on this. - The issue of when to use storing or non-storing modes was also highlighted and agreed that further study and potential recommendations would be of value. 3) The Minimum Rank Objective Function with Hysteresis (Phil Levis) draft-ietf-roll-minrank-hysteresis-of-02 - Final discussion focused on Parent set, including Parent set size and configuration parameters. - Co-chairs felt the document was stable and any further comments can be addressed during WG LC. 4) Reactive Discovery of Point-to-Point Routes in Low Power and Lossy Networks (Emmanuel) draft-ietf-roll-p2p-rpl-02 - Co-chairs clarification on the term “Temporary DAG” used in draft, in the RPL specification (version 19) it is termed “Local DAG”. - Next steps include identifying independent implementations and potential interoperability testing. - There is a dependency on this document with RPL. It uses common messages, including DIO, and framework around core specification. - Co-chairs reminded WG and authors that they should reuse components of RPL as possible. The mechanisms defined in this document will then build on the core specification to provide P2P enhancements. 5) "Selective DIS for RPL" draft-dejean-roll-selective-dis (Dominique Barthel) - In the event of a dense network how a collision mitigation option has been proposed. The author would like to ascertain if the requirement needs to be addressed and if the proposed solution is acceptable. The format is described in final slide set (backup section). 6) "RPL Objective Function 0" draft-ietf-roll-of0-07 (Pascal) - Request from Pascal that implementers of the draft provide feedback and best practice suggestions on some of the options in the draft. - Co-chairs highlighted that Phil Levis had an outstanding question on the draft and they requested that he reviews the new text and confirm his comment has been addressed. 7) Lightweight Key Establishment and Management Protocol in Dynamic Sensor Networks (KEMP) (Qui Ying) draft-qiu-roll-secure-router-00 - Co-chairs conducted a poll on who had read the draft, no hands raised. 8) The Direction Field in Routing Metric/Constraint Objects Used in RPL (Emmanuel) draft-goyal-roll-metrics-direction-00 - No comments. 9) A Mechanism to Measure the Quality of a Point-to-point Route in a Low Power and Lossy Network (Emmanuel) draft-goyal-roll-p2p-measurement-01 - No comments. - Chairs will poll the list for WG adoption 10) Identifying Defunct DAGs in RPL (Emmanuel - 5mn) [100] draft-goyal-roll-defunct-dags-00 General comment that “Academics” need to get out more to understand the operational aspect of RPL and how people implement and deploy. ☺ 11) Version Number Authentication and Local Key Agreement (Amit Dvir) draft-dvir-roll-security-extensions-00 - Jabber question “How can we assume the node cannot be compromised within time T after its deployment?” Amit mentioned that the scheme assumes it cannot be compromised as in practice its a few minutes or even seconds, so an observer can check it is not compromised during start-up. - The next version of the draft the authors will include an authentication mechanism. - Co-chairs polled the room; three people have read the draft. 12) ROLL Re-chartering Discussion (All) - It was mentioned that an Abstract Management Model (protocol neutral) document might beneficial for exploring how to manage an LLN. The document would detail components (node, router, link, network wide) of a LLN that would need to be managed. - Additionally a document could be created to explore requirements and objectives for deterministic routing within an LLN. This document would investigate the applicability of existing source-based routing and label switching techniques that may be applicable to LLNs based on requirements. - JP asked T. Clausen if he wanted to add more comments since he mentioned on Jabber that he could not make his comments. T. Clausen's reply: No.