Hi Jouni, all, to complement what has been written already, I think once we agreed on a framework it helps doing a comprehensive gap analysis. Based on a common understanding of how DMM scenarios work, the framework defines functions that can add to a basic set of assumed mobility/tunnel management functions. If a certain mobility protocol supports one or more DMM function intrinsically, fine. If not, the available protocol may be extended to enable that function. Or an external protocol can be considered to provide that functionality.
My impression from past DMM meetings is that there is a common understanding of DMM in general, but folks still have different expectation from DMM support and associated IP address continuity. Also assumptions on hosts are different (e.g. multi-anchor capability, maintenance of associated uplink routes, etc.). A framework could classify DMM functions as mandatory or optional, dependent on the target scenario and level of path optimization that should be supported. Some functions according to draft-liebsch-dmm-framework-analysis could be placed even on the host to see if host mobility protocols can contribute to DMM operation. I see the framework as required portion of the top-down gap analysis and design approach. Last but not least it allows building a DMM architecture and deployment of associated protocols beyond mobility protocol scope, e.g. by contributions from the routing plane above mobility anchor level. I understand that the DMM WG may focus mainly on mobility protocols first, but the framework can document DMM solutions in a broader scope, which may attract readers beyond the DMM WG. marco >-----Original Message----- >From: dmm-boun...@ietf.org [mailto:dmm-boun...@ietf.org] On Behalf Of >Jouni Korhonen >Sent: Donnerstag, 20. Dezember 2012 14:50 >To: dmm@ietf.org >Subject: [DMM] DMM and the framework discussion > >Folks, > >During the last meeting we had a presentation(s) on DMM framework >approaches. The discussion was somewhat left unfinished during and after the >meeting. It might be useful to have the discussion on the mailing list now and >not to postpone in to the next f2f meeting :) > >Have a read e.g. on draft-liebsch-dmm-framework-analysis and say what you >have to say about DMM frameworks. I keep hearing we lack the framework and >at least I need to understand the possible need for one.. > >- Jouni >_______________________________________________ >dmm mailing list >dmm@ietf.org >https://www.ietf.org/mailman/listinfo/dmm _______________________________________________ dmm mailing list dmm@ietf.org https://www.ietf.org/mailman/listinfo/dmm
