MULTIMOB Notes (Thanks to Akbar Rahman) Chairs Administrativia - Carlos - Asked that all drafts be submitted according to IETF cutoff submission guidelines in the future - Bechet - Gave status of WG drafts. Said that Hitoshi's draft would soon enter WGLC Hitoshi WG draft update - Hitoshi - Gave summary of changes. No major changes. - Hitoshi - Gave a summary of various timer values for wireless links - There were various comments/questions during the presentation including the following: o Stig - Noted that the query interval numbers have small variation and asked if they shouldn't be larger variation? o Bechet - Asked which value was for WiFi and which was for cellular o Thomas - Asked whether the robustness variable should be increased rather than decreased? o Stig - Said that he agreed with Thomas' suggestion for increasing robustness variable - Hitoshi - Said he will follow up on all the questions above and asked for people's opinions on the mailing list. - Bechet - Re-affirmed that he would like the I-D to go to WGLC after this meeting Shuai Source mobility draft - Shuai - This is a merged draft with him and Thomas - Shuai - Gave description of proposal and discussed some open issues - There were various comments/questions during the presentation including the following: o Jeffrey - PIM WG is considering moving the DR which may help solve one of the open issues o Hitoshi - Asked for clarification about MAG acting as PIM router vs DR o Thomas - Said that several deployment scenarios are being covered and one has MAG acting as DR o Sri - Asked why MAG could not do forwarding in the experiments that Shuai described o Hitoshi - Mentioned that there is no source-stop scheme o Stig - In PIM-SM there is no source-stop problem - Shuai - Said he will follow on any required updates to the questions above - Bechet - Thanked authors for merging their drafts. Will ask for mailing list comments and will ask for WG adoption on mailing list Carlos Handover I Carlos - Gave summary of status, changes, motivation and advantages - There were various comments/questions during the presentation including the following: o Thomas - Is multicast handover handled independently of unicast handover? o Thomas - Asked for clarification about reactive mode handover mechanism? o Marco - Why do you not have MAGs transferring context? What is the need for synchronizing multicast and unicast handovers? o Hitoshi - What is difference between proactive and reactive handovers? o Sri - Asked to clarify L2 triggers and use - Carlos - Said he will follow on any required updates to the questions above Thomas Handover II - Thomas - Gave an overview of the flow, design objectives, and various questions asked on the mailing list - There were various comments/questions during the presentation including the following: o Carlos - Are MN modifications within the scope of the charter? o Marco - Asked what direction of transfer gave the most advantages? - Thomas - Said he will follow up on any required updates to the questions above Marco Issues in Fast Handover Solutions - Marco - Gave input to the discussion towards finding a good choice for fast handover solutions - There were various comments/questions during the presentation including the following: o Thomas - Did not agree with the mechanism illustrated for buffer management in the network - Marco - Said he will follow up on any required updates to the questions above - Bechet - Asked Marco for a list of thing that we can agree on and to post that on the list Hitoshi Tunnel convergence draft - Hitoshi - Gave background, changes, - There were various comments/questions during the presentation including the following: o Thomas - asked what was being copied to the MRIB? You cannot express PMIP routing in one MRIB. o Sri - Proxy MIP routing is policy based routing o Stig - Does not see how to do policy based routing with PIM o Bechet - This issue seems open in the draft (i.e. PMIP vis PIM routing) o Seil - May be difficult for roaming case. PIM-SM may not be supported by all operators. What about the operator not deploying PIM-SM? Through several meeting and mailing discussions, we've known that no solutions can satisfy all the requirements. We need to consider PIM-SM case "as one of them" but tunnel convergence solution should not be dependant on the algorithm used in specific multicast routing protocol. - Hitoshi - Said he will follow up on any required updates to the questions above Juan Carlos Providing multicast service draft - Juan Carlos - Gave overview of the merger of two existing drafts (Direct routing and MTMA concepts) - There were various comments/questions during the presentation including the following: o Thomas - Did you solve how to get the multicast group identification issue o Bechet - This draft solves the tunnel convergence. This draft is ready to become WG draft. o Stig - Ask on the mailing list to give comments and in one month to ask for WG adoption.