Jouni, >-----Original Message----- >From: Jouni Korhonen [mailto:[email protected]] >Sent: Mittwoch, 28. Mai 2014 19:55 >To: Marco Liebsch; [email protected] >Subject: Re: [DMM] rechartering > >Marco, > > >5/28/2014 3:35 PM, Marco Liebsch kirjoitti: >> Hi Jouni, all, >> >> let me pick up some comments again, which we should clarify. >> >> About anchor selection: >> >> Referring to the milestones, having the anchor re-selection document >> separated from the advanced anchor selection only by 3 months does not >> seem to be useful, even if you consider re-selection to be used only >> for advanced DMM deployment and operation, whereas basic anchor >> selection is required for any DMM deployment with decentralized data plane >anchors. >> >> I understand this document treats solely selection and runtime >> re-selection, not the associated establishment of routing/tunnel >> states according to the > >Actually, I was in an opinion that re-selection would also cover the needed >parts >for possible re-routing of traffic.. whatever the mechanisms is.
Ok, if we are clear about this, I am fine. To confirm your understanding: the milestone about anchor selection is solely about the selection process, while the milestone about anchor re-selection is about runtime selection of a new anchor plus traffic routing. Correct? Then I still believe it would make sense to cover that in a single document. If we want to separate documents, then we may address anchor selection and runtime re-selection (just the selection) in a single document, as the same algorithms and protocol components should apply to find a suitable anchor, whereas a separate document treats the traffic routing to the selected anchor. No too strong opinion here and no push to change wording as long as the listed milestones allow all required work to be done. > >Does anyone else share the same view? > >> selected anchor. So, I think both aspects, selection and re-selection, >> should go into a single document. Unless the WG decides that >> re-selection is too advanced, then it may go into a separate document >> when the base selection spec is almost ready, as it will depend on it. >> >> About forwarding path and signaling management: >> >> In particular if the support of runtime/mid-session anchor change is >> considered for the first set of DMM solution documents, the described >> work item about 'forwarding path and signaling management' should be >> given a milestone, as this is essential if anchors are changed >> mid-session and IP address continuation is expected. So far there is >> no document about this work item in the charter list. > >Right, any suggestions for the actual document name and time lines? That one I understand is associated with the traffic routing in the above documents, no? marco > >> Inter-working between mobility functions and network nodes, e.g. >> routers, requires the use of some non-mobility protocol. DMM should at >> least describe which states (identifiers, locator addresses/IDs, ..) >> to expose through these protocols. That's what I understood behind >> mobility state exposure, whereas the text in the work items list reads >> more like this is about announcements from the network to an MN about >> the characteristics of an IP address to support the MN in picking the >> right address for an application. I know it's more a terms issue, but we need >clarification what we expect from a work item. > >The latter is what I had in mind.. mainly the MN<->network communication. > >> Hope the above makes sense and we can easily clarify. > >Sure. > >- Jouni > > >> >> marco >> >> >> >> >>> -----Original Message----- >>> From: dmm [mailto:[email protected]] On Behalf Of Jouni Korhonen >>> Sent: Dienstag, 27. Mai 2014 10:51 >>> To: [email protected] >>> Subject: Re: [DMM] rechartering >>> >>> >>> Now that the gap analysis I-D is almost on the stage of leaving the >>> WG and the requirements I-D has almost completed IESG, it would be >>> time to return to the rechartering topic. >>> >>> First, the latest revision can be found at: >>> https://github.com/jounikor/dmm-re-charter >>> >>> Second, have a look at it. There are few changes proposed by Alper >>> eons ago and corrected milestones pointed by Behcet. >>> >>> Third, let us get this finally done.. >>> >>> - Jouni >>> >>> 4/22/2014 3:55 PM, Jouni Korhonen kirjoitti: >>>> Folks, >>>> >>>> Sorry for letting this topic to rot in a dark for the couple of last >>>> weeks. I'll crank out a revision shortly.. >>>> >>>> - Jouni & Dapeng >>> >>> _______________________________________________ >>> dmm mailing list >>> [email protected] >>> https://www.ietf.org/mailman/listinfo/dmm _______________________________________________ dmm mailing list [email protected] https://www.ietf.org/mailman/listinfo/dmm
