IETF
dmm@jabber.ietf.org
Friday, November 8, 2013< ^ >
Room Configuration
Room Occupants

GMT+0
[02:52:41] Peer Azmat joins the room
[02:52:45] Peer Azmat leaves the room
[04:39:16] Peer Azmat joins the room
[04:39:26] Peer Azmat leaves the room
[05:16:52] Peer Azmat joins the room
[05:25:13] Peer Azmat leaves the room
[16:06:10] Peter McCann joins the room
[16:45:04] Paolo Saviano joins the room
[17:01:31] vvucetic joins the room
[17:02:14] Peter McCann 2 joins the room
[17:04:44] tobia joins the room
[17:05:40] sarikaya2012 joins the room
[17:07:38] Alex Petrescu joins the room
[17:08:03] Brian Haberman joins the room
[17:08:08] <Alex Petrescu> slide: Agenda
[17:08:43] <Alex Petrescu> Anthony Chan is AC approaching
[17:09:33] <Alex Petrescu> slide: Issue tracker status
[17:10:07] <Paolo Saviano> Slide 1: DMM Requirements
[17:10:07] <Alex Petrescu> slide: Version 09 (Septemner 28 2013)
[17:10:24] <Alex Petrescu> slide: Version 10 (November...
[17:10:27] <Alex Petrescu> Slide: next step
[17:10:31] <Paolo Saviano> Slide 5: Next step
[17:10:41] <Alex Petrescu> Sri Gundaveli i sSg
[17:10:43] <Paolo Saviano> Slide 4: Version 10 (November 7 2013)
[17:10:45] <Paolo Saviano> Slide 3: Version 09 (September 28 2013)
[17:10:57] <Alex Petrescu> SG: not really a chance I had to read the doucment, I quickly browsed, maybe some editorial I have issues
[17:11:02] <Alex Petrescu> SG: no technical issues I have
[17:11:18] <Alex Petrescu> SG: I need a way how ot provide those comments, some secitons require more work
[17:11:25] <Alex Petrescu> SG: I think how can we make progress?
[17:11:32] <Alex Petrescu> SG: I can take more time, send more comments
[17:11:38] <Alex Petrescu> AC: can you send by deadline?
[17:11:41] <Alex Petrescu> AC: when?
[17:11:47] <Alex Petrescu> AC: I am ok but when?
[17:11:51] <Alex Petrescu> SG: a week
[17:12:05] <Alex Petrescu> Jouni Korhonen is JK: four months week?
[17:12:11] <Alex Petrescu> SG: next week is 3GPP
[17:12:23] <Alex Petrescu> Kostas Pentikousis is KP
[17:12:31] <Alex Petrescu> KP: maybe we need step forwad with SG
[17:12:36] <Paolo Saviano> Presentation stopped
[17:12:41] <Alex Petrescu> SJK: he has clear depending issues , that we relate him
[17:12:54] <Alex Petrescu> JK: answer Sri today
[17:13:12] <Alex Petrescu> JK: long week
[17:13:28] <Alex Petrescu> Juan Carlos Zuniga is JZ
[17:13:33] <Paolo Saviano> Slide 1: Distributed Mobility Management: Current
[17:13:33] <Alex Petrescu> JZ presenting
[17:13:46] <Paolo Saviano> Current presenter: Juan Carlos Zuniga
[17:13:47] <Paolo Saviano> Slide 1: Distributed Mobility Management: Current
[17:13:49] <Paolo Saviano> Slide 2: Current Status
[17:13:51] Peer Azmat joins the room
[17:15:00] <Paolo Saviano> FYI, a Meetecho session is available at  www.meetecho.com/ietf88/dmm
[17:15:33] <Paolo Saviano> Slide 38: The New Section 5
[17:15:44] <Alex Petrescu> slide 38??? :-)
[17:15:49] <Paolo Saviano> Slide 39: Gap Summary 1/2
[17:16:30] <Paolo Saviano> yes :)
[17:16:40] <Paolo Saviano> Slide 40: Gap Summary 2/2
[17:17:20] <Paolo Saviano> Slide 41: Next Steps
[17:17:30] <Alex Petrescu> Alper Yegin is AY approaching
[17:17:58] <Paolo Saviano> Slide 39: Gap Summary 1/2
[17:18:19] <Alex Petrescu> AY: anchor selection needed HSS? changing the anchor duruing an IP session ongoing is going to break it, unless solution I dont foresee?
[17:18:27] <Alex Petrescu> AY: You can _no_ chang ein the middle
[17:18:31] <Alex Petrescu> JZ: this is issue
[17:18:38] <Alex Petrescu> JZ: we highlight the gap
[17:18:51] <Alex Petrescu> AY: not somehting you try to do?
[17:18:54] <Alex Petrescu> JZ: we can try
[17:18:59] <Alex Petrescu> AY: dont know how yoiu do that?
[17:19:06] <Alex Petrescu> AY: is this suggesting a solution?
[17:19:11] <Alex Petrescu> AY: who says we need to do thaT?
[17:19:24] <Alex Petrescu> JZ: some demos we ad we were able... assign new anchor
[17:19:33] <Alex Petrescu> JZ:; but in that case... there is app layer, and TCP layer
[17:19:41] <Alex Petrescu> that was AY
[17:19:46] <Alex Petrescu> AY: two different IP sessions?
[17:20:09] <Alex Petrescu> AY: session continuity for the video, but tow distinct IP sessions we had, we didnt change the anchor, but the IP session we changed
[17:20:20] <Alex Petrescu> AY: this text suggests somehting which was not ion the demo
[17:20:27] <Alex Petrescu> AY: IP session? higher layer sessio,?
[17:20:30] <Peter McCann> I think the key is to un-couple the mobility from the change of IP address.
[17:20:32] <Alex Petrescu> JK: youll hear Sri
[17:20:40] <Alex Petrescu> SG: injecting routes in the BGP
[17:20:54] <Alex Petrescu> SG: redundancy, you can relocate a funciton East-West, multiple scenarios
[17:21:01] <Alex Petrescu> SG: if user moving, anchor closer to North
[17:21:15] <Alex Petrescu> SG: separate control plane, inject routes in data plane, Ryuji's draft(!)
[17:21:38] <Alex Petrescu> AY: BGP route update is not really 'anchor', they have routing state
[17:21:44] <Alex Petrescu> AY:... thats an anchor (LMA, etc)
[17:21:52] <Alex Petrescu> AY: BGP is not anchor, i slike routing reflector
[17:21:59] <Alex Petrescu> SG: control plane to a different anchor
[17:22:19] <Alex Petrescu> SG: every time we move mobility anchor, we anchor sessio, we need to insert respective routing table into
[17:22:20] <Peter McCann> Some have defined "anchor" as topological destination of an address.  By that definition, BGP route updates change the anchor.
[17:22:32] <Alex Petrescu> Dave Moses is DM
[17:22:46] <Alex Petrescu> DM: optimal initial anchor we provide?  I dont think we do, and its not always optimal
[17:22:57] <Alex Petrescu> KP: existing solutions a tbest can provide
[17:23:03] <Alex Petrescu> DM: why not just say initial?
[17:23:24] <Alex Petrescu> JK: read archive, paste on some, methodical knowledge we can anchor 6463 (RFC) in the PMIP space
[17:23:35] <Alex Petrescu> JK: not optimal, but somehow slected
[17:23:41] <Alex Petrescu> DM: its not optimal
[17:23:46] <Alex Petrescu> Georgios Karagiannis is GK
[17:23:57] <Alex Petrescu> GK: another issue need consideration is traffic steering
[17:24:12] <Alex Petrescu> GK: suppose select a new anchor and traffic actually going to previous anchor has to be redirected
[17:24:21] <Alex Petrescu> GK: somehow need solution for redirecting traffic
[17:24:42] <Alex Petrescu> GK: the other point is - where do we start redirecting - is it the previous point, or up in the infra?
[17:24:53] <Alex Petrescu> GK: another q is for how long do we keep that redirection?
[17:25:04] <Alex Petrescu> JS: steering traffic - you mean w/o experiencing mobility?
[17:25:11] <Alex Petrescu> GK: _w/_ mobility
[17:25:27] <Alex Petrescu> GK: the goal is to change the anchor pointa nd take care the traffic from previous to the new one
[17:25:33] <Alex Petrescu> Dapeng Liu i sDL
[17:25:43] <Alex Petrescu> DL: do you think Global HAHA can do change?
[17:25:47] <Alex Petrescu> AY: antying?
[17:25:52] <Alex Petrescu> AY: I have to look at details
[17:26:03] <Alex Petrescu> JK: this steers some discussion
[17:26:10] <Alex Petrescu> JK: people broght additional things?
[17:26:24] <Alex Petrescu> JK: anyone disagrees there is a gap what we had today what people want to do?
[17:26:29] <Peter McCann> We need to agree on a definition of "anchor"
[17:26:35] <Peter McCann> Alex, can you relay?
[17:26:52] <Alex Petrescu> JK: if you think there is absolutely no gap, or if you want to do more intelligent anchor seleciton - anyone thing there is ... gap there?
[17:27:59] <Peter McCann> That definition assumes we are using tunnels.
[17:28:04] <Alex Petrescu> JK: the anchor is the one which terminates tunnel terminates address
[17:29:20] <Alex Petrescu> JK:...
[17:29:23] <Peter McCann> Is a BGP update a change of anchor?
[17:29:32] <Alex Petrescu> SG: in netext work we already split CP/DP (control plane data plane) work
[17:29:42] <Alex Petrescu> SG: we have ability to register IP address CP and DP
[17:29:48] <Alex Petrescu> SG: single box DP/CP
[17:29:58] <Alex Petrescu> SG: in on eline card DP, in another line CP
[17:30:19] <Alex Petrescu> SG: already ack CP/DP can be separated from protocol perspective
[17:30:21] <Alex Petrescu> JK: yes
[17:30:27] <Alex Petrescu> SG: can we keep the reception you say?
[17:30:28] <Peter McCann> I think anchor point needs to be defined with respect to the data plane.
[17:30:56] <Alex Petrescu> We need someone to relay Pete's comments.  I tend to agree with them.
[17:31:18] <Alex Petrescu> JZ: this work is current practice we compare gaps we want to achieve
[17:31:28] <Alex Petrescu> JZ: not existing
[17:31:48] <Alex Petrescu> JZ: come back to original q frol Jouni I think we identify a few gaps an dI dont think current solutions address them
[17:32:00] <Paolo Saviano> Slide 40: Gap Summary 2/2
[17:32:03] <Alex Petrescu> JZ: not yet feedback I have, but new revision has already enough material
[17:32:15] <Alex Petrescu> JK: onc enew version we have around we issue tracker tickets
[17:32:38] <Alex Petrescu> JK: anyone has anything has to say about correspondence
[17:32:43] <Alex Petrescu> JK: be quick with comments
[17:32:48] <Alex Petrescu> AY: deadline for that?
[17:33:08] <Alex Petrescu> JK already said wants something before ?(in relationship with) London
[17:33:12] Peer Azmat Shah joins the room
[17:33:31] <Alex Petrescu> DL: we discuss recharter, the gap analysis document should be aligned with this document
[17:33:45] <Alex Petrescu> DL: the gap analysis document should justify this; maybe we can take a look later.
[17:33:52] <Alex Petrescu> Alper Yegin is AY going to present
[17:34:03] <Paolo Saviano> Presentation stopped
[17:34:17] <Alex Petrescu> (earlie it was Danny Moses DM an dnot David Moses)
[17:34:33] <Paolo Saviano> Slide 1: On    
 Demand    
 Mobility    
 Managemen
[17:34:38] <Paolo Saviano> Current presenter: Alper
[17:34:39] <Paolo Saviano> Slide 1: On    
 Demand    
 Mobility    
 Managemen
[17:34:42] <Paolo Saviano> Slide 2: Mobile    
 IP
[17:35:46] <Paolo Saviano> Slide 3: Approach
[17:35:58] <Paolo Saviano> Slide 4: (with    
 respect    
 to    
 IP-­‐lay
[17:38:20] <Paolo Saviano> Slide 5: RFC    
 5014
[17:38:54] tobia leaves the room
[17:40:49] <Paolo Saviano> Slide 6: SoluRon
[17:41:29] <Brian Haberman> What exactly is an Unanchored Address?  Did Alper describe that?
[17:41:43] <Paolo Saviano> Slide 7: Policy
[17:41:45] <Peter McCann> It is tied to the current access point.
[17:41:54] <Peter McCann> It goes away on any mobility event.
[17:42:11] <Brian Haberman> @Pete: I thought that was the Access Address.
[17:42:31] <Peter McCann> Access anchored can be maintained across mobility events within the same access network.
[17:42:39] <Alex Petrescu> Charles PErkins is CP approaching mic, also SG
[17:42:40] <Brian Haberman> Ah, right.
[17:42:45] <Paolo Saviano> Slide 8: QuesRons    
 and    
 comments?
[17:42:50] <Alex Petrescu> SG: go back to slide x
[17:42:52] <Paolo Saviano> Slide 4: (with    
 respect    
 to    
 IP-­‐lay
[17:42:59] <Alex Petrescu> SG: what is the def of Access Network?
[17:43:06] <Brian Haberman> @Pete: Thanks.
[17:43:08] <Alex Petrescu> SG: that is where the node attaches to the Internet
[17:43:21] <Alex Petrescu> SG: multilayer hierarchical net, essential home addresses, how are you mapping that?
[17:43:43] <Alex Petrescu> SG: access network address and ... address is the sale, my CoA; but there is also the local address, w/o mobility
[17:43:53] <Alex Petrescu> SG: with this new definition, what use?
[17:44:00] <Alex Petrescu> AY: you mean context of HMIP?
[17:44:04] <Alex Petrescu> AY: regional CoA?
[17:44:12] <Alex Petrescu> SG: clear, so far, we have just  2 defs
[17:44:20] <Alex Petrescu> SG: that is something... now you split
[17:44:27] <Alex Petrescu> access network address and anchor address
[17:44:34] <Alex Petrescu> AY: anchored address? you mean?
[17:44:45] <Alex Petrescu> AY: even if I change my AR then I can still keep using it
[17:45:00] <Alex Petrescu> AY: allocate the ... address frim an HMAP or from a default access router?
[17:45:06] <Alex Petrescu> AY: ... as opposed to ...
[17:45:15] <Alex Petrescu> AY: ..HA.. wiht respect to
[17:45:19] <Alex Petrescu> SG: centrally locateD?
[17:45:37] <Alex Petrescu> SG: multil layer mobility hierarchy; essentially Two-level hierarchy, but which address is which?
[17:45:51] <Alex Petrescu> AY: if home address stays same all the time, that is ... anchoring
[17:45:58] <Alex Petrescu> AY: home, stays home all the time
[17:46:03] <Alex Petrescu> AY: fixed IP address
[17:46:07] <Alex Petrescu> SG: take off-line
[17:46:41] <Alex Petrescu> CP: 1st I basically support what you described, 2nd I probably eli,inate need for later presentaitons in the session, because ther eis strong overlap with draft liu Mobility API
[17:46:49] <Alex Petrescu> CP: but maybe we can find a way to work together
[17:46:53] <Alex Petrescu> AY: indeed
[17:47:04] <Alex Petrescu> DL: granularity is whaT? per-prefix? or per-package?
[17:47:11] <Alex Petrescu> AY: it's actually per-flow
[17:47:22] <Alex Petrescu> AY: each time you bind;, there is a flow, so it's per flow
[17:47:31] <Alex Petrescu> DL: how about multiple appli connecitons?
[17:47:39] <Alex Petrescu> AY: each flow different flg, yes
[17:47:43] <Alex Petrescu> DL: is that possible?
[17:47:54] <Alex Petrescu> DL: for one work package there is one flow, not sure
[17:48:14] <Alex Petrescu> AY: twp sockets, on each socket you set flags accordingly, each socket would go to a different IP address
[17:48:21] <Alex Petrescu> DL: relax the routing mechanism in the Host?
[17:48:31] <Alex Petrescu> AY: not thinking so, we can discuss, just trat them differently
[17:48:35] <Alex Petrescu> Marco Liebsch i sML
[17:48:52] <Alex Petrescu> ML: home address is what? how many addresses? Two-types of addresses?
[17:49:04] <Alex Petrescu> ML: by this type, not sure Mobile Node is in position to decide what is doable
[17:49:24] <vvucetic> please tell the speakers NOT to touch mike when they talk
[17:49:29] <Alex Petrescu> AY: app is aware of this tbale: if I need this reachbaility then I need that address; if not needs it session continuity then choose that address
[17:49:35] <Alex Petrescu> ML: so that's topologic
[17:49:37] <Alex Petrescu> AY: yes
[17:49:38] Peer Azmat leaves the room
[17:49:50] <Alex Petrescu> Ryuji Wakikawa is RW
[17:49:58] <Alex Petrescu> RS: clash up between entries in the table?
[17:50:01] <Alex Petrescu> RW: all the times?
[17:50:09] <Alex Petrescu> RW: turn off again, different address?
[17:50:14] <Alex Petrescu> AY: HA may give same I paddress
[17:50:22] <Alex Petrescu> RW: if so then tsame?
[17:50:38] <Alex Petrescu> AY: if you want same address through reboots, then you'd need a home network address
[17:50:40] Peer Azmat Shah leaves the room
[17:50:45] <Alex Petrescu> RW: thi sis like definition
[17:50:47] <Alex Petrescu> AY: yes
[17:51:02] <Alex Petrescu> RW: second, app point of view - what is the benefit between access network and address
[17:51:07] <Alex Petrescu> AY: low latency yuou get
[17:51:13] <Alex Petrescu> RW: but you get address assignment
[17:51:20] <Alex Petrescu> RW: too complicatd may be for app
[17:51:22] <Alex Petrescu> AY: hmmm
[17:51:27] <Alex Petrescu> RW: explain the fourm
[17:51:37] Lorenzo Miniero joins the room
[17:51:39] <Alex Petrescu> AY: that table supposes to be everything you ever need
[17:51:44] <Alex Petrescu> AY: if you need latency
[17:51:49] <Alex Petrescu> (forum===coloring)
[17:52:27] <Alex Petrescu> JZ: in the meeting MIF archi design team we discussed whqt to do API and characteristics like this, maybe in the future we might be good idea also show it in MIF
[17:52:30] <Alex Petrescu> JK: good point
[17:52:31] <Paolo Saviano> Presentation stopped
[17:52:44] <Lorenzo Miniero> Slide 1: DMM Framework based on Functional Elemen
[17:52:49] <Lorenzo Miniero> Current presenter: Marco Liebsch
[17:52:50] <Alex Petrescu> (I wanted to relay message but not sure when to do it)
[17:52:50] <Lorenzo Miniero> Slide 1: DMM Framework based on Functional Elemen
[17:52:51] <Lorenzo Miniero> Slide 2: Preamble..
[17:54:27] <Lorenzo Miniero> Slide 3: Methodology
[17:55:17] Peer Azmat joins the room
[17:55:33] <Lorenzo Miniero> Slide 4: Methodology
[17:56:44] <Lorenzo Miniero> Slide 5: One abstract deployment model
[17:57:40] <Lorenzo Miniero> Slide 6: Another abstract deployment model
[17:58:24] <Lorenzo Miniero> Slide 7: Summary .. What’s Next?
[17:59:19] <Alex Petrescu> JL: do I need an assumption of the solution
[17:59:28] <Alex Petrescu> ML: any kind of DMMM
[17:59:35] <Alex Petrescu> ML: we received som ecomments from Carlos
[17:59:49] <Alex Petrescu> ML: but the fw does not mandate any
[17:59:51] <Alex Petrescu> DL: other potential solutions may fit in this fw?
[18:00:07] <Alex Petrescu> ML: yes, that's why,... model i susin gPMIP, MIP
[18:00:23] <Lorenzo Miniero> Presentation stopped
[18:00:33] <Alex Petrescu> JK: we will see this fw things later on (fw==framework)
[18:00:43] <Lorenzo Miniero> Slide 1: Distributed Mobility  Management Fram
[18:00:44] <Alex Petrescu> H. Anthony Chan is AC is presenting
[18:00:50] <Lorenzo Miniero> Current presenter: Anthony Chan
[18:00:51] <Lorenzo Miniero> Slide 1: Distributed Mobility  Management Fram
[18:01:02] <Lorenzo Miniero> Slide 2: DMM framework
[18:01:04] <Lorenzo Miniero> Slide 3: Unified formulation of Internet mobi
[18:01:52] <Lorenzo Miniero> Slide 4: Existing protocol: MIPv6
[18:02:10] <Lorenzo Miniero> Slide 5: Existing protocol: PMIPv6
[18:02:15] <Lorenzo Miniero> Slide 6: MIPv6/PMIPv6
[18:02:27] <Lorenzo Miniero> Slide 7: Host­ vs Network­based mobility ma
[18:02:36] <Lorenzo Miniero> Slide 8: Hierarchical
[18:02:57] <Lorenzo Miniero> Slide 9: Deploying MM in each network
[18:03:02] <Lorenzo Miniero> Slide 10: Selective mobility support
[18:03:10] <Lorenzo Miniero> Slide 11: Selective mobility support
[18:03:16] <Lorenzo Miniero> Slide 10: Selective mobility support
[18:03:50] <Lorenzo Miniero> Slide 11: Selective mobility support
[18:04:15] <Lorenzo Miniero> Slide 12: Selective mobility support
[18:04:19] <Lorenzo Miniero> Slide 13: Distributed mobility anchors­Architec
[18:04:23] <Lorenzo Miniero> Slide 14: DMM
[18:05:26] <Lorenzo Miniero> Slide 15: Comparing framework against  DMM req
[18:05:42] <Alex Petrescu> DL: can you justify why do we need this fw?  Can you find any gaps?
[18:06:04] <Alex Petrescu> AC: you can do find gaps, and there are other places, other protocols - any venodr/architecture have some of the...
[18:06:07] <Alex Petrescu> KP of EICT
[18:06:19] <Alex Petrescu> KP: had we already these questions again and again
[18:06:28] <Alex Petrescu> KP: but we need a way to talk these things in a higher layer?
[18:06:47] <Alex Petrescu> KP: there is a tendency in the IETF mobility work to do particular extensions and completely loose capacity to see higher
[18:07:05] <Alex Petrescu> KP: if we continue so, very soon we will miss that train as well (SDN) because of this ptptpt
[18:07:12] <Alex Petrescu> KP: we miss problem statement
[18:07:19] <Alex Petrescu> KP: we have reqs for too long
[18:07:27] <Alex Petrescu> KP: not a q if whether we need a fw
[18:07:37] <Alex Petrescu> KP: we need a tool to think in higher level terms
[18:07:49] <Alex Petrescu> KP:... and advance faster with the popular way of work here
[18:07:57] <Alex Petrescu> KP: need to adopt this and move fwd.
[18:08:08] <Alex Petrescu> KP: this is work can happen simultaneoulssy with other things around us
[18:08:13] <Alex Petrescu> JK: I dont think...
[18:08:26] <Alex Petrescu> KP: strong believer in at least the fw is already capturing details of other things
[18:08:40] <Alex Petrescu> KP: obvously fw is not detailed, it should capture the essential
[18:08:58] <Alex Petrescu> KP: the cost of context switch between each of these slots will be so high, noit productive
[18:09:06] <Alex Petrescu> KP: 2 things orthogonal an docmplimentary
[18:09:14] <Alex Petrescu> KP: not sure we need to finish the fw before..
[18:09:23] <Alex Petrescu> KP: think different, more futuristic
[18:09:35] <Alex Petrescu> KP: 5G projects already EU
[18:09:40] <Alex Petrescu> KP: CfP I do
[18:09:54] <Alex Petrescu> (CfP is Call for Papers, EU i sEuropean Union)
[18:10:04] <Lorenzo Miniero> Presentation stopped
[18:10:23] <Lorenzo Miniero> Slide 1: draft-xiong-dmm-ip-reachability-00
[18:10:32] <Lorenzo Miniero> Current presenter: Chunshan Xiong
[18:10:32] <Lorenzo Miniero> Slide 1: draft-xiong-dmm-ip-reachability-00
[18:10:44] <Alex Petrescu> Xiong chunshan is XC (Sam?) is presenting
[18:10:57] <Lorenzo Miniero> Slide 2: MN IP Reachability Problem
[18:11:24] <Lorenzo Miniero> Slide 3: DDNS Method
[18:12:42] <Lorenzo Miniero> Slide 4: Server Registration Method
[18:14:33] <Lorenzo Miniero> Slide 5: iste Serv r with M ice ID
[18:15:10] <Alex Petrescu> " iste Serv r with M ice ID"? funny
[18:15:25] <Lorenzo Miniero> Slide 6: iste Serv r with M ice ID
[18:15:41] <Lorenzo Miniero> ops sorry, these slide titles didn't parse well :)
[18:15:55] <Lorenzo Miniero> Slide 6: Server Registration Method: Server Central Mode
[18:16:40] <Lorenzo Miniero> Slide 7: Server Registration Method: Server Combined Mode
[18:17:19] <Lorenzo Miniero> Slide 8: Issues to be investigated
[18:18:27] <Lorenzo Miniero> Slide 9: How to Forward
[18:19:30] <Alex Petrescu> DL: group, understanding motivaiton, what scenario IP reachability is useful?
[18:19:49] <Alex Petrescu> XC: both conceptually, I hear there are a lot of people with these comments in previous meetings
[18:20:15] <Alex Petrescu> XC: IP reachability is somewhat different than routing, diff mechanisms - eg old address use tunnel to transform tunnel
[18:20:26] <Alex Petrescu> XC: I Prelated to routing orgainzation issues
[18:20:58] <Alex Petrescu> CP: briefly, proposal wher some ppplicaitons use server to provide app an address endpooint, others use DNS, so you make this app-specific for end 2 end connectivity
[18:21:09] <Alex Petrescu> CP: but since it does it anyway, then.. is this correct.?
[18:21:32] <Alex Petrescu> XC: intention is not to provide detail solution for IP attachmend - intention when you have current solution use it, and
[18:22:07] <Alex Petrescu> XC: and if thsis paper through we find out, where to support this IP mobile ?  intention is not to define new mechanism.  Just find out what are the issues with reachbaility;
[18:22:09] <Alex Petrescu> JK: ok thank you
[18:22:27] <Alex Petrescu> JK: Ryuji if here? (RW)
[18:22:35] <Lorenzo Miniero> Presentation stopped
[18:22:36] <Alex Petrescu> RW approach presentatin floor
[18:22:49] <Lorenzo Miniero> Slide 1: Stateless user­plane architecture fo
[18:22:56] <Lorenzo Miniero> Current presenter: Ryuji Wakikawa
[18:22:57] <Lorenzo Miniero> Slide 1: Stateless user­plane architecture fo
[18:22:58] <Lorenzo Miniero> Slide 2: Splitting Control and User Plane
[18:24:15] <Lorenzo Miniero> Slide 3: Simple Configuration
[18:24:35] <Lorenzo Miniero> Slide 4: Stateless user­plane architecture  f
[18:26:19] <Alex Petrescu> AY: interesting approahc
[18:26:25] <Alex Petrescu> AY: need see performance indications
[18:26:53] <Alex Petrescu> AY: when I look I wonder BGP scalable?  load on BGP to serve a typical mobnetwork, and convergence rate?  Data to present so people comfortable?
[18:27:05] <Alex Petrescu> AY: roaming in 3rd nets where no same BGP cloud?
[18:27:24] <Alex Petrescu> RW: perf - any volunteer I really happy to work with, otherwise self, but some university stundet?
[18:27:30] <Alex Petrescu> RW: data next IETf
[18:27:43] <Alex Petrescu> RW: some qualification when mobile moves 3rd party
[18:27:48] <Alex Petrescu> Brian Haberman is BH
[18:27:56] <Alex Petrescu> BH: single AS?  or propagation outsidE?
[18:27:59] <Alex Petrescu> RW: no, single
[18:28:02] <Alex Petrescu> BH: good thanks
[18:28:14] <Alex Petrescu> DL: stateless you mean?
[18:28:18] <Alex Petrescu> RW: you mean this edge router, no need t
[18:28:23] <Alex Petrescu> RW: this packet has no state
[18:28:44] <Alex Petrescu> JK: you said you already work on this?  you have required some draft you/elsebody?
[18:28:55] <Alex Petrescu> RW: this extension eeded;, new type ID in BGP
[18:29:02] <Alex Petrescu> JK: already have you a proposal?
[18:29:10] <Alex Petrescu> RW: yes
[18:29:13] <Alex Petrescu> x: I coauthored
[18:29:21] <Alex Petrescu> AY: there is state - the BGP routes
[18:29:24] <Alex Petrescu> RW: route can be state
[18:29:37] <Alex Petrescu> RW: but from this guy, when moving, this router can not distinguish
[18:29:40] <Alex Petrescu> RW: thats why
[18:29:49] <Alex Petrescu> Behcet Sarikaya is BS
[18:29:54] <Alex Petrescu> BS: also tunnel is state
[18:30:04] <Alex Petrescu> ML: argument scalability DMM emipies
[18:30:13] <Alex Petrescu> ML: per host states are transit
[18:30:22] <Alex Petrescu> ML: longer you maintain longer you have to...
[18:30:26] <Alex Petrescu> ML: interesting to see
[18:30:35] <Alex Petrescu> JK: you already proposal have?  BGP type extension?
[18:30:42] <Alex Petrescu> JK: you need potentially?
[18:30:52] <Alex Petrescu> RW: like a guideline to combine mutiple protocols
[18:30:59] <Alex Petrescu> JK: loo k fo rhome for it?
[18:31:04] <Alex Petrescu> RW: ho^pefully DMM
[18:31:11] <Alex Petrescu> John...
[18:31:19] <Alex Petrescu> J: stateless, chargin how would you do then?
[18:31:22] <Alex Petrescu> RW: good q
[18:31:28] <Peter McCann> John Kaippallimalil
[18:31:32] <Alex Petrescu> RW: NESISIS chargin for this
[18:31:39] <Alex Petrescu> RW:that introduces extra state
[18:31:49] <Alex Petrescu> SG: youre piece a function in the .. element?
[18:32:01] <Alex Petrescu> SG: w/o tunnel, piece of functions appky?
[18:32:10] <Alex Petrescu> SG: parameters, use that to apply all policies functions?
[18:32:18] <Alex Petrescu> SG: IP packet, I paddress, 2 diff enterprises
[18:32:35] <Alex Petrescu> RW: DPIL charging, generaliwing services, put into data planes, or just dont need accounting
[18:32:51] <Alex Petrescu> Carlos Bernardos is CB
[18:33:01] <Alex Petrescu> CB: any differences too many open sources?
[18:33:07] <Alex Petrescu> ...
[18:33:13] <Alex Petrescu> CB: for the minute taker
[18:33:15] <Lorenzo Miniero> Presentation stopped
[18:33:20] <Alex Petrescu> CB presenting
[18:33:26] <Lorenzo Miniero> Slide 1: DMM Open Source platform
[18:33:31] <Lorenzo Miniero> Current presenter: Carlos Bernardos
[18:33:32] <Lorenzo Miniero> Slide 1: DMM Open Source platform
[18:33:33] <Lorenzo Miniero> Slide 2: Demos @ IETF 83
[18:33:36] <Lorenzo Miniero> Slide 3: Open Source platform
[18:33:55] <Lorenzo Miniero> Slide 4: Open Source platform
[18:34:11] <Alex Petrescu> (is 'MIPS' an architecture? or unit?)
[18:34:34] <Peter McCann> MIPS is a CPU architecture.
[18:34:50] <Lorenzo Miniero> Presentation stopped
[18:35:20] <Alex Petrescu> JK: help Chairs and Ad, discuss, based on gap analysis, feeling of people
[18:35:31] <Alex Petrescu> JK: should lead this work in coming years or months
[18:35:42] <Alex Petrescu> JK: today no fixed structure what to discuss today
[18:35:50] <Alex Petrescu> JK: not proposal charter do we have
[18:35:56] <Alex Petrescu> JK: reaching end of current charter
[18:36:09] <Alex Petrescu> JK: strong statement is ther ein Charter: either identify or you close the group
[18:36:25] <Alex Petrescu> JK: hope to have continue, approve before London meeting
[18:36:33] <Alex Petrescu> JK: depends on couple documents been in and out
[18:36:53] <Alex Petrescu> JK: I state existing WG docs, when get they go in a state reqs leads the group
[18:37:14] <Alex Petrescu> JK: gap analysis, state call for last call, more or less ready start working on next steps
[18:37:34] <Alex Petrescu> JK: I bring the kind of first step, slide on gap analysis
[18:37:40] <Alex Petrescu> JK: exa,ple things we identified
[18:37:51] <Alex Petrescu> JK: good indication of couple things could work on
[18:37:56] <Alex Petrescu> JK: open mic
[18:38:02] <Lorenzo Miniero> Slide 39: Gap Summary 1/2
[18:38:04] <Lorenzo Miniero> Current presenter: Chairs
[18:38:05] <Lorenzo Miniero> Slide 39: Gap Summary 1/2
[18:38:09] <Alex Petrescu> JK: there is material to send off line
[18:38:14] <Alex Petrescu> JK: so offline gorup can select that
[18:38:22] <Alex Petrescu> JK: present offline group conclusion or so
[18:38:34] <Alex Petrescu> K: expectin ghuge queue
[18:38:39] <Lorenzo Miniero> Slide 40: Gap Summary 2/2
[18:38:44] <Alex Petrescu> KP: only because you ask
[18:38:53] <Alex Petrescu> KP: it would be catastrophoic to close the WG
[18:39:10] <Alex Petrescu> KP: this is the only future oriented WG, with all due respect to other WGs looking at mobility
[18:39:22] <Alex Petrescu> KP: yes kind of some time, we could do thing sfaster
[18:39:34] <Alex Petrescu> KP: bunch EU projects H2020 will feed lots of material
[18:39:42] <Alex Petrescu> KP: not sure close now and then open up another?
[18:40:03] <Alex Petrescu> KP: current analysis from it, review indicates even if we look at more futuristic 2020, as is
[18:40:24] <Alex Petrescu> KP: we have been having corridor discussion, because SDN is taking over, people talk virtualiwinzg everythin
[18:41:03] <Alex Petrescu> KP: still think firm believer is that through valley of gaps, centralized SDN kind of stuff; by that time we could have viable distirbuted solution through traditional IP solutions
[18:41:14] <Alex Petrescu> KP: not as glorious as SDN, it would save the day time on the road one, current hype
[18:41:21] <Alex Petrescu> KP: lots people good ideas need a home
[18:41:36] <Alex Petrescu> KP: if by chance we decide no progress I'd feel sad IETF no future-priented WG
[18:41:40] <Alex Petrescu> KP: 2050
[18:41:44] Michal Wodczak joins the room
[18:42:04] <Alex Petrescu> KP: downside/up no question about closing WG; we should try to do more than pointing extensions
[18:42:12] <Alex Petrescu> KP: step back we need; we have intellect do better
[18:42:25] <Alex Petrescu> KP: 3GPP slow anyway
[18:42:32] <Alex Petrescu> GP: if we try, then maybe improitant
[18:42:39] <Alex Petrescu> CP: agree with just said
[18:42:54] <Alex Petrescu> CP: seems strong interest in 3GP EPC archi, nlack eye
[18:43:14] <Alex Petrescu> CP: IETF in unique position, except, tehre is bit of ... management issue 3GPP
[18:43:33] <Alex Petrescu> CP: not closing group down, but enlarging Charter, find out how to get IETF protocols adopted by 3GPP
[18:43:50] <Alex Petrescu> CP: dot 11 carrying more license than 3GPP, in several years well see
[18:43:59] <Alex Petrescu> CP: terrible if we get
[18:44:12] <Alex Petrescu> CP: get more and more data wireless techs not 3GPP, because business didnt adopt IETF protocols
[18:44:29] <Alex Petrescu> CP, 3rdly, we went through gap analysis, discussions, rectancgles, - where else is this done in IETF?
[18:44:37] <Alex Petrescu> CP: whole set of rectangles in SDN groups?
[18:44:52] <Alex Petrescu> CP: not here to provide leadership if not, then things wrongly done would be at IETF
[18:45:06] <Alex Petrescu> JW: in general, I agree, work done here is good, we agree d on reqs
[18:45:10] <Alex Petrescu> JZ is that
[18:45:25] Michal Wodczak leaves the room
[18:45:31] <Alex Petrescu> JZ: allow for room for little mor einnovative soluions; work is happening in different areas
[18:45:52] <Alex Petrescu> JZ: next week IEEE, WiFi, encourage groups to be opening up ifaces to be dynamically configurables
[18:46:01] <Alex Petrescu> JZ: encourage group push same time
[18:46:18] <Alex Petrescu> JZ: more about defining in such a way that we allow ... as well as enhancements
[18:46:26] <Alex Petrescu> JZ: othe rinitiatives we should allow
[18:46:35] Lorenzo Miniero leaves the room
[18:46:39] <Alex Petrescu> BH: w/o hat on
[18:46:43] <Alex Petrescu> (so I dont type :-)
[18:46:50] <Alex Petrescu> BH: part of that is
[18:47:02] <Alex Petrescu> BH: the various mobility protocols developped havent got as much traction we hoped
[18:47:10] <Alex Petrescu> BH: doesnt mean moving forward
[18:47:21] <Alex Petrescu> BH: care should we not rearchitecting selves again and again
[18:47:23] <Alex Petrescu> AY: for you
[18:47:25] <Alex Petrescu> BH: no
[18:47:31] <Alex Petrescu> BH: should make you happy
[18:47:45] <Alex Petrescu> BH: if you have a proposal you like, by all means bring it to Charter, show
[18:47:57] <Alex Petrescu> BH: as long as you hsow up with credible evidence people want to use
[18:47:59] <Alex Petrescu> BH: eg
[18:48:08] <Alex Petrescu> BH: WLAN mobility protocol? bring a charter and people
[18:48:14] <Alex Petrescu> BH: relates to this group?
[18:48:23] Lorenzo Miniero joins the room
[18:48:35] <Alex Petrescu> BH: we have a charter to this group; if we let recharter, they tend to walk off start do things elswehere
[18:48:47] <Alex Petrescu> BH: need focus for deliverables, then we can have discussion what to do in this WG
[18:48:55] <Alex Petrescu> BH: if you have new things, then bring me charter
[18:49:07] <Alex Petrescu> BH: JC I encourage mobility space people get involved in IEEE
[18:49:16] <Alex Petrescu> BH: IEEE interesting set of problems they do
[18:49:25] <Alex Petrescu> CP: IEEE is a big place
[18:49:31] <Alex Petrescu> CP: you mean 802.11?
[18:49:40] <Alex Petrescu> CP: even there, how many task groups there, 15 or so
[18:49:43] <Alex Petrescu> CP: specific?
[18:49:48] <Alex Petrescu> CP: only RAN? MIH?
[18:50:00] <Alex Petrescu> CP: whereas I dont disagree what you say, but what speciifc?
[18:50:03] <Alex Petrescu> BH: OmniRAN
[18:50:12] <Alex Petrescu> CP: problematic because separate
[18:50:17] <Alex Petrescu> BH: no I cant do about that
[18:50:22] <Alex Petrescu> CP: believe me I understand
[18:50:34] <Alex Petrescu> SG: offline we discussed
[18:50:40] <Alex Petrescu> SG: chairs open that slide
[18:50:48] <Alex Petrescu> slide "Proposed Extensions"
[18:51:02] <Alex Petrescu> SG: how operators can truly deploye network, what do they call distributed?
[18:51:15] <Alex Petrescu> SG: every ufdntcion, CP/DP
[18:51:22] <Alex Petrescu> SG: that seems to be _the_ ...
[18:51:28] <Alex Petrescu> SG: if you agree aiwth that,
[18:51:37] <Alex Petrescu> SG: look at Ryuji's presentation, that's wha the wants
[18:51:42] <Alex Petrescu> SG: other
[18:51:56] <Alex Petrescu> SG: if we can solve somle problems, maybe ND extensions, maybe steer traffic
[18:52:01] <Alex Petrescu> SG: Ryuji presenter
[18:52:09] <Alex Petrescu> SG: IoT applications, state applications
[18:52:14] <Alex Petrescu> SG: not as far as SDN do I go
[18:52:25] <Alex Petrescu> SG: minor protocol tweaks and realize architectural model good
[18:52:44] <Alex Petrescu> SG: essentially - how do you deploy a network?  then we can extend all protoocls, 3GPP, RSGPP
[18:52:55] <Alex Petrescu> SG: can we do some work in that spac eplease?
[18:52:59] <Alex Petrescu> SG: at lease we have a list
[18:53:12] <Alex Petrescu> DL: the DMM module, any relationship with the fw in this droup discussed?
[18:53:17] <Alex Petrescu> SG: not me looked at it, but Marco
[18:53:27] <Alex Petrescu> SG: I put this in very generic very high level, for any operator
[18:53:34] <Alex Petrescu> SG: considerations for DMM...
[18:53:40] <Alex Petrescu> SG: distirbute it how do I?
[18:53:45] <Alex Petrescu> SG: should I maybe aggregate?
[18:53:49] <Alex Petrescu> SG: high level
[18:53:57] <Alex Petrescu> KP: dnt think this is the case
[18:54:12] <Alex Petrescu> KP: I agree with Brian, need to go to reqs and gap analysis
[18:54:19] <Alex Petrescu> KP: maybe some people have fit?
[18:54:29] <Alex Petrescu> DL: is this just justified by the gap analysis?
[18:54:36] <Alex Petrescu> SG: in lrge part yes, almost everything
[18:54:44] <Alex Petrescu> KB is ... vfrom Verizon
[18:54:59] <Alex Petrescu> KB: there is in SDN different ways of implementing, GW, MME, PCRF
[18:55:12] <Alex Petrescu> KB: look t virtualization
[18:55:25] <Alex Petrescu> KB: if hrdware blade I have then I can have in one, and nin another one have
[18:55:38] Michal Wodczak joins the room
[18:55:43] <Alex Petrescu> KB: if MME is on same blade or chassis, then we dont need these extending interfaces
[18:55:56] <Alex Petrescu> KB: maybe this is what you DMM guys have, maybe extned, maybe not
[18:56:18] <Alex Petrescu> KB: I know yo ufocus on distirbute, decentralized, but I also think many people work in virtual VPC, different models there are
[18:56:23] <Alex Petrescu> KB: interfaces disappear
[18:56:28] <Alex Petrescu> KB: then what?
[18:56:38] <Alex Petrescu> KP: but virtual EPS already defined is
[18:56:41] Michal Wodczak leaves the room
[18:56:42] <Alex Petrescu> EPC is that
[18:56:48] Michal Wodczak joins the room
[18:56:51] <Alex Petrescu> KP: you ar ein FORCES WG?
[18:56:55] <Alex Petrescu> KB: yes
[18:57:08] <Alex Petrescu> KP: you dont define anything new
[18:57:24] <Alex Petrescu> KB: Tzo things there are, interfaces disappear what does it mean to mobility
[18:57:32] <Alex Petrescu> KB: virtualiwation is...
[18:57:40] <Alex Petrescu> KP: in the SDNRG we have a temrinology draft
[18:57:49] <Alex Petrescu> KP: when colocated we call it a protocol
[18:58:04] <Alex Petrescu> JK: ok , thank you, thank you
[18:58:04] <Alex Petrescu> AY: been there forever
[18:58:13] <Alex Petrescu> AY: state of WG ways forward
[18:58:21] <Alex Petrescu> AY: good people and good ideas we are but
[18:58:23] sarikaya2012 leaves the room
[18:58:32] <Alex Petrescu> AY: energy level is low because we stuck in reqs
[18:58:44] <Alex Petrescu> AY: not suggest skip that, but boos energy level, otherwise
[18:58:51] <Alex Petrescu> AY: I suggest
[18:59:08] <Alex Petrescu> AY: we had presentaitons about solutions but IMHO not enough discussion had we on tht
[18:59:08] Peer Azmat leaves the room
[18:59:14] <Alex Petrescu> AY: but people still interest in that?
[18:59:29] <Alex Petrescu> AY: spend more time on discussion solutions I suggest, not to build solutions, but energy level
[18:59:48] <Alex Petrescu> AY: suggest to next IETf we can have conferenc calles, people presnet their ideas, propoer technicla discussions
[19:00:18] <Alex Petrescu> AY: between now and next IETF
[19:00:36] <Alex Petrescu> KP: we can hos tmeeting in Berling
[19:00:36] <Alex Petrescu> AY: second is 3GPP
[19:00:36] <Alex Petrescu> HD: proposal I heard
[19:00:47] <Alex Petrescu> HD: we could focus on the protocol here
[19:00:57] <Alex Petrescu> BH: there are guidelines to do interim meetings
[19:01:08] <Alex Petrescu> BH: if you can make progress then I will ... you.
[19:01:16] <Alex Petrescu> JK: keep discussion ongoing
[19:01:25] Brian Haberman leaves the room
[19:01:30] <Alex Petrescu> JK: please be open with the discussion, so people outside see that we want to do stuff
[19:01:35] Lorenzo Miniero leaves the room
[19:01:37] <Alex Petrescu> JK: not everybody is part of secrecy and emails
[19:01:40] <Alex Petrescu> JK: thank you
[19:01:52] <Alex Petrescu> JK: hope to see before London, telephones whatever, thanks.
[19:02:01] <Alex Petrescu> DL: bluesheets?
[19:02:06] <Alex Petrescu> Adjourned it seems to be.
[19:02:13] Alex Petrescu leaves the room
[19:02:51] vvucetic leaves the room
[19:02:51] Michal Wodczak leaves the room
[19:02:56] Peter McCann 2 leaves the room
[19:03:03] Peter McCann leaves the room
[19:03:23] Paolo Saviano leaves the room
[19:32:42] Peer Azmat joins the room
[19:36:18] Peer Azmat leaves the room
[20:26:34] Peer Azmat joins the room
[20:29:54] Peer Azmat leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!