Behcet, On Tue, Jun 30, 2015 at 1:24 AM, Behcet Sarikaya <sarikaya2...@gmail.com> wrote:
-- snip -- > Wait a minute, there is more. > It seems like you have not read Sections 4.2 and 4.3. > Yes, XMPP and OpenFlow are for Section 4.1 for switch control on > mobile backhaul. > This is related to a university project we were involved which used > public domain OpenFlow tools like OpenDayLight. > > In route establishment, I am using Netconf or RPC protocol which you > missed big and that was one of the main additions in Rev. 02. > Please check it again. > Wow, thanks. I had overlooked it. I had jumped to conclusions when I read section 4.1 that cover not only L2 but also L3 hand-over. I couldn't imagine that each layer has different technique for hand-over. Why is that? I want to clarify one thing on your drat. Sec 4.3 in the draft mentions that i2rs agent as netconf client but it is opposite as far as I know. Am I wrong? > > > But what I mentioned FPCP in vEPC is as a way to export mobility info to > BGP > > speaker > > from control-plane nodes of mobility management. > > > > And more, many parameters should be aligned for these nodes to utilizing > > data-plane > > network so that the config data model should also be defined in FPCP. I > > think that's not > > part of signaling protocol roles. > > > > As I said in my mail before, I wonder why we need all those things? > Why do we need another client/server system maybe other than what i2rs > is defining? > I think this is a big question FPCP lovers should answer? My point is > that at least you should not need it in vEPC, I am saying that as > someone who supports vEPC :-) > AFAIK, FPCP design is alined with i2rs architecture so that it has the data-model with yang schema based on its information model of data-plane node abstraction (Marco's idea) but currently it's an appendix. The information model would make clear what are the mobility specific things that might augment i2rs model. If nothing, there's no need to define it, yes. As a co-author of FPCP, you are welcome to review the parameters which FPCP defined whether each of those are really needed or not. > > > Thought? > > Another main addition is Rev. 02 was Section 5 on multicast support > which you missed or not commented? > Cool. Does mobile node really need multicast? Is there any document that justify for that? cheers, --satoru
_______________________________________________ dmm mailing list dmm@ietf.org https://www.ietf.org/mailman/listinfo/dmm
