[05:22:43] Meetecho joins the room
[05:25:08] Lorenzo Miniero joins the room
[05:25:09] Igor Bryskin joins the room
[05:26:03] danielking.net joins the room
[05:26:38] Srihari Ramachandra joins the room
[05:29:20] adrianfarrel joins the room
[05:30:04] Lou Berger joins the room
[05:30:24] Lou Berger has set the subject to: IETF106
[05:31:34] Giridhar Mandyam joins the room
[05:33:13] Giridhar Mandyam leaves the room
[05:34:18] Srihari Ramachandra leaves the room
[05:34:52] Srihari Ramachandra joins the room
[05:35:57] Suneesh Babu joins the room
[05:36:29] Suneesh Babu leaves the room
[05:38:48] Daniele Ceccarelli joins the room
[05:39:35] Fatai Zhang joins the room
[05:41:26] Simon Pietro Romano joins the room
[05:46:29] Thomas Peterson joins the room
[05:55:30] Vishnu Ram joins the room
[05:59:23] Thomas Peterson leaves the room
[05:59:44] Thomas Peterson joins the room
[06:16:06] Bernard Sales joins the room
[06:31:56] Michael Scharf joins the room
[06:44:53] <Lou Berger> @daniele will you be at the second session?
[06:44:53] Thomas Peterson leaves the room
[06:45:46] <Daniele Ceccarelli> yes
[06:45:47] <adrianfarrel> @lou, I will, if you want to swamp me out
[06:46:31] <Daniele Ceccarelli> however my presentation will be given by Julien
[06:46:47] <Daniele Ceccarelli> he's there in person
[06:47:54] danielking.net leaves the room
[06:48:03] <Daniele Ceccarelli> Julien Maisonneuve (not Meuric)
[06:49:00] <Lou Berger> great - thank you
[06:49:40] <Lou Berger> will do the ETSI in the next sessuon
[06:49:58] <Daniele Ceccarelli> ok
[06:50:11] <Daniele Ceccarelli> it'll come after another cups of coffee, it's even better
[06:50:15] <Lou Berger> actually, might be able to squeeze it in if adrian is fast
[06:50:24] <Lou Berger> ;-)
[06:50:28] <Lou Berger> I need one twpo
[06:50:42] <Lou Berger> too!
[06:57:45] danielking.net joins the room
[06:58:23] <Daniele Ceccarelli> can you please ask how IETFers can contribute?
[06:58:44] <Daniele Ceccarelli> i do it offline via my company represnetatives in ETSI but not anyone can use a channel like that
[07:00:08] danielking.net leaves the room
[07:00:59] <Daniele Ceccarelli> no there is no formal releationship AFAIK
[07:03:29] <Lou Berger> we're out of time
[07:03:36] wang jun joins the room
[07:03:45] <Daniele Ceccarelli> ok, i'll comment here: this has NOTHING to deal with network slicing :)
[07:04:11] <Lou Berger> will repeate that
[07:04:19] <Daniele Ceccarelli> thx
[07:04:49] <Daniele Ceccarelli> bye
[07:05:04] Meetecho leaves the room
[07:05:11] Lou Berger leaves the room
[07:05:11] Lorenzo Miniero leaves the room
[07:05:15] Fatai Zhang leaves the room
[07:05:15] Srihari Ramachandra leaves the room
[07:05:15] Daniele Ceccarelli leaves the room
[07:05:15] Igor Bryskin leaves the room
[07:05:15] Simon Pietro Romano leaves the room
[07:05:15] wang jun leaves the room
[07:05:15] Michael Scharf leaves the room
[07:05:15] Vishnu Ram leaves the room
[07:05:16] Bernard Sales leaves the room
[07:08:22] VirtualQueue_eZUtsmzJ leaves the room
[07:19:11] adrianfarrel leaves the room
[07:41:52] adrianfarrel joins the room
[07:42:02] adrianfarrel leaves the room
[07:48:53] Martin Horneffer joins the room
[07:57:33] Martin Horneffer leaves the room
[08:56:55] Meetecho joins the room
[09:05:15] Bernard Sales joins the room
[09:09:05] Srihari Ramachandra joins the room
[09:09:24] Lou Berger joins the room
[09:10:33] adrianfarrel joins the room
[09:14:56] Michael Scharf joins the room
[09:16:36] Daniele Ceccarelli joins the room
[09:17:20] Michael Scharf leaves the room
[09:17:44] zhenqiang li joins the room
[09:28:44] Simon Pietro Romano joins the room
[09:30:19] <Daniele Ceccarelli> network slicing might be just a use case but it shows up 5 times in a single slide
[09:30:47] <zhenqiang li> that's cool
[09:30:48] <Lou Berger> do you want me to repeat this?
[09:31:02] <Lou Berger> note this is a different draft
[09:31:20] <Lou Berger> the VPN+ one is a WG doc, this is not
[09:31:58] <Daniele Ceccarelli> in a better shape: please make sure that network slicing is just one of the use cases of this work and not trying to do what is in the scope of the DT team
[09:33:02] <Lou Berger> In queue after me
[09:35:20] Fatai Zhang joins the room
[09:36:10] <Lou Berger> @meetecho remote volume is low
[09:36:22] <Meetecho> Lou Berger: ack
[09:37:44] <Daniele Ceccarelli> the confusion in this document makes me feel there is the same confusion in the VPN+ framework (confusion in terms of scope). Please make sure this is not another attempt to define network slicing but focus on VPN+ and, if possible, not just on deterministic VPNs but VPNs that can deliver traffic with given KPIs/SLAs
[09:39:01] <Lou Berger> how do thouse two things differ?
[09:40:04] <Daniele Ceccarelli> deterministic VPN makes me think of packets delivered in order, not a single packect dropper. a VPN with constraints is a VPN that might have upper bound limints in term of delay or have a given amount of reserved bandwidht in the network
[09:40:23] <Daniele Ceccarelli> (sorry for typos...trying to type as fast as possilbe)
[09:47:53] danielking.net joins the room
[09:58:16] <adrianfarrel> As a contributor to the VPN+ framework, I'm a bit downcast by your comments, Daniele. The term "deterministic VPN" is yours and the only mention of "deterministic" in the draft is in the context of DetNet.
[09:59:59] <adrianfarrel> Obviously, it's a WG draft and can be modified as the WG wants, but at the time it was adopted it contained some clear text about slicing and was (IMHO) TEAS putting a stake in the ground about slicing. Indeed, it includes the text "Note that, in this document, the four terms "VPN", "Enhanced VPN" (or "VPN+"), "Virtual Network (VN)", and "Network Slice" may be  considered as describing similar concepts dependent on the viewpoint from which they are used."
[10:00:47] <Daniele Ceccarelli> i was referring to a previous comment talking about Network Slicing vs Deterministic VPNs. I felt the need to clarifiy "not just" deterministic VPNs
[10:02:40] <adrianfarrel> In other words, in *my* mind there is not a significant difference between VPN+ and network slicing. The main thing was that the term "network slicing" was clearly poisoned, and it was useful to use "VPN+" because when looked at from outside, they are the same
[10:02:50] <Daniele Ceccarelli> Indeed my concerns are more against the presentation than against document. I don't have major issues with the document, but the presenation seems to be very much network slicing oriented.
[10:04:11] <Daniele Ceccarelli> but if VPN+ and Network Slicing are the same thing why are duplicating the work in the DT?
[10:04:34] <adrianfarrel> I didn't set up the DT after the WG had adopted the framework :-)
[10:04:40] zhenqiang li leaves the room
[10:04:53] <Daniele Ceccarelli> this is another issue :)
[10:06:28] <Daniele Ceccarelli> still in your mind (since we're running a parallel session) what is the difference between VPN+ and service to tunnel binding?
[10:06:44] <Daniele Ceccarelli> isn't the final goal the same?
[10:07:38] <adrianfarrel> I wonder whether one of the hidden issues here is that a "VPN" is not a technology, it is a service. The mismatch on that thought bit us doing L3SM, and perhaps it is biting here as well. There is a difference between the service (whether you call it an enhanced VPN or a network slice or a virtual network it is essentially a similar/same service), and the delivery mechanisms (control plane and data plane which may also be similar for the two concepts).
[10:08:03] <adrianfarrel> And, yes, service to tunnel binding is also relevant
[10:08:46] <adrianfarrel> The framework was *intended* (may not have achieved!) to set the scene to developing technology solutions
[10:09:30] <adrianfarrel> I certainly hope the DT is looking at the VPN+ framework to see in what ways what it says is different from their view of network slicing
[10:09:53] <adrianfarrel> Perhaps, also, compare and contrast with ACTN VN
[10:09:59] <Daniele Ceccarelli> this could save Jari a lot of headhaches :)
[10:10:25] <adrianfarrel> Well yes. Before starting research read the literature?
[10:14:17] Michael Scharf joins the room
[10:15:24] David Sinicrope joins the room
[10:26:25] <David Sinicrope> Hi Adrian - the enhanced vpn draft needs some work. It confuses network slicing with the VPN service you note and then mixes in a good dose of the need for VPN/TE binding
[10:27:55] <David Sinicrope> It also make statements about providing levels of traffic isolation that have little to do with the VPN or the binding, but the underlying technology used to provide determinism
[10:29:20] <David Sinicrope> ... and that was only section 1. ;-)
[10:35:30] <adrianfarrel> Thanks, Dave, you make my point
[10:36:00] <adrianfarrel> You are talking about VPN as though it is a technology not a service
[10:36:40] <adrianfarrel> The request for isolation is a service request.
[10:36:55] <adrianfarrel> The delivery of isolation is a technology.
[10:37:23] <adrianfarrel> The request of r a slice or an enhanced VPN is a service request
[10:37:38] <David Sinicrope> Not my words, I'm regurgitating the draft
[10:37:42] <adrianfarrel> How you achieve those things is technology-specific
[10:38:04] <David Sinicrope> I agree with what you are saying, but that's not what I'm seeing in the draft
[10:38:29] <adrianfarrel> Then we should look through it together, because I look at the same words and see something different :-)
[10:39:23] <adrianfarrel> For example, you use the word "deterministic" but where do you find that in the draft that is not about DetNet?
[10:39:26] <David Sinicrope> "Unlike a traditional    VPN, an enhanced VPN can achieve greater isolation with strict    performance guarantees. "
[10:39:55] <David Sinicrope> as you say, its not the VPN that gives the guarantee but the technology underneath
[10:40:33] <adrianfarrel> Mutter: it's not the connection service that gives the guarantee of connectivity, it's the technology underneath
[10:40:34] <David Sinicrope> the words I used were "deterministic transport" which is a general reference to DetNet, TSN and FlexE
[10:41:03] <David Sinicrope> all of which are mentioned in the draft
[10:41:35] Srihari Ramachandra leaves the room
[10:41:43] <adrianfarrel> yes they are, in the sections that provide a list of possible candidate technologies
[10:42:28] <adrianfarrel> But recall that when you describe a slice you are describing an offering
[10:42:46] <adrianfarrel> If, in your quote, it said "offer" instead of "achieve"... ?
[10:43:03] danielking.net leaves the room
[10:43:08] adrianfarrel leaves the room
[10:43:14] Bernard Sales leaves the room
[10:43:14] David Sinicrope leaves the room
[10:43:14] Michael Scharf leaves the room
[10:43:14] Daniele Ceccarelli leaves the room
[10:43:14] Fatai Zhang leaves the room
[10:43:15] Simon Pietro Romano leaves the room
[10:44:12] Lou Berger leaves the room
[10:44:22] Meetecho leaves the room
[14:55:11] Lou Berger joins the room
[23:35:09] Lou Berger leaves the room