 Welcome to the panel discussion on the Cloud Infrastructure Telecom Task Force, CNTT, from zero to first deliverable in three months. We're going to be talking about how we were able to start up a project so quickly. And we're still going strong a year later. So with that, I would like to open by introducing our panelists. So I'm going to go around very quick introductions. Scott. Hi, I'm Scott. I'm a lead architect at AT&T and the interim TSC chairperson for CNTT. I'm Gerge. Hi, I'm working as a senior open source specialist for Nokia. And I'm mostly active in the RA2 in CNTT, a bit in the reference model also. And Tom. Hi, I'm Tom. I'm working in GSME as technical director and helping to work CNTT and to allow the transition of the CNTT reference model to GSME specification. And finally, but last but not least, Mark. Hello, everyone. I am Mark Cohen from Spirant. I've been involved in Linux Foundation networking activities over the past decade. And I'm involved in the verification and validation activities within the reference conformance project. So first, I'm going to get into, so how did this thing get started anyhow? So I am actually the person on the panel who's been with the project the earliest. I actually joined in December. I was not part of the GSME project, which started in October. But there was a separate project that happened around the same time. And then at March, in March at the ONES meeting in San Jose, we had an informal gathering where we realized there were two projects that were both working on the same gap, perceived gap in the infrastructure support of network function virtualization, i.e. for telecoms who need to have infrastructure to support their virtualized products or projects or systems, whatever. So we got together a very small group of telecom people. In fact, it was 12 people met in Barcelona to really ideate what this was going to look at. And that included somebody from GSME and also a person from the LFN, as well as somebody from the OpenStack Foundation. And we all agreed that it was very important to include not only the telecoms, but also the vendors and the testing community and the community at large. So the big kickoff was in July in Paris on record-breaking heat. And we then, coming out of the Paris event, we aimed at getting our first release out in September. And we were, in fact, successful. And we have since had several other releases. So here is a picture of all of us at the Paris event inside where it was slightly cooler. So anyone, I guess, Gergé, you were there. What was your, and, Mark, I think you were there as well. So what were your thinking? Yeah, that's a very good question. So there was this legend among the vendors that operators started some activity, but first vendors were not invited. So we didn't know what is this whole thing about. And just before this July 2019 event, we received a PDF document, which was the CNTD specification at that time. And we could comment that. And I think that was a good start of the discussion. I think that was also a very important step to involve the vendors into this discussion and open up the space for collaboration. And Tom, although I know you weren't there, it was very important that it was a joint effort between GSMA and the Linux Foundation. So I think if you would like to make a few comments about that, why that's so important. I wasn't there, so I'm really having to go by hearsay. But obviously, common infrastructures, different profiles are very important for our members. So as GSMA is really happy to support this. There's also my colleague Mikaeli is on the picture, so he was there. But yeah, I think this is a very important activity for us to make sure that the network infrastructure of operators can be deployed as efficiently as possible. And I said, we're happy to support this. And we will continue to support it going forward as well. And I just want to comment that this is a somewhat old slide. I think it's not covering everybody. But as you can see, there's been a huge outpouring of support right from the beginning. And I will want to remark that the Barcelona event, several of the vendors did expressed a little bit of concern that they weren't invited to the Barcelona event. But I will point out there were only 12 people total. So it was. And during that event, we decided that we needed to open it up to the vendors. It was actually a very critical decision, I think. So with that, I would like to move on to a discussion about what we are actually all about. And I would like to turn this over to Scott to go into the details. Great. Thanks, Beth. So we're currently at our release coming up in September for the Barak feature set. And on the screen, we see a couple of rows. We've got the RM, which is the reference model. That's really the overall guiding set of scope for CNTT. And you'll notice the one after the rest of the first row. And then you'll see a two on the bottom. It's loosely broken up into two parts, a one stream, we call it, and a two stream. The one stream is all based around OpenStack or VM-based. Then the two stream is based around containers, Kubernetes as the platform. So we've got a lot of good topics that independently are released as a part of the Barak feature set in September of this year. And I think the overriding theme you'll see throughout this is a lot about containers. And you'll see it in the top row, it's sitting bottom row. The one stream with OpenStack continues to release based on the major OpenStack releases, continuing to add more details, cookbooks on potential ways to install it and get it set up and also conformance test suite. The two stream, the RA2 is the requirements that describe how an NFVI will be standing after it's been built, along with a conformance suite test that and a sample implementation, the RA2, so reference implementation. So those are really just kicking off, and we'll have a good suite of requirements and documentation coming forth in the September release there. So Gerge, I know you've been very active in this release. Love to hear from comments from you. Yeah, so I'm mostly focusing into the RA2, so the container or Kubernetes-based reference architecture. And in RA2, we are now doing a rework of chapter 4. Chapter 4 is the chapter where we define the different components and the configuration of the different components of a Kubernetes-based cloud infrastructure. And the aim of this change is to be able to trace back the requirement. And Mark, I'd like also for you to make a comment about the certification program. Of course, that's a very important part of this whole project. Yes, and chapter 8 actually addresses the certification. But if you'll notice, there isn't a lot of detail there at this point because we've been focusing on actually the reference model and the reference implementation. And certification becomes particularly important as we begin to commercialize CNTT. And both operators as well as vendors are thinking about how they can build confidence among end users as well as the overall CNTT community. And one way to do that is to ensure that we have a rigorous testing and certification activity to validate all the activities that we have, or at least the technical requirements that we have underway. And you can expect to see additional efforts along these lines over the next number of months. Of course, the Linux Foundation networking has been involved in certification for a while. And we do have the OPNFE verification program, or OVP, which is going to be instrumental. And of course, the OPNFE platform is going to be critical as we repurpose that to be used for CNTT validation and testing activities. And Scott, I'd like to have you add some color around. What do you think were the key elements that made this project so far successful, I should say? I think, really, it's a strong collaboration. That's the biggest thing here. It's open source community. I think it's got a wide variety of companies that are involved. And listening to everyone speak come together. Sometimes we all disagree on various topics, but yet we all come together in the end because we know this is important to the industry as a whole. And what you're seeing on the screen here is a testament to that. This is last week's status at our governor's call. It shows contributions from week after week. We're in a season in some parts of the world of vacationing. So we have quite a variety. But normally, the contributions are very strong. We work in GitHub, pull requests, issues are well worked. But plenty of back and forth going on until we come to a consensus. And we can merge back into the main branch. Tom, you want to comment on the strength of the community here? Yes, I've not been involved in all activities, but I've been involved mainly in the reference mode, I'll touch on why that is later on. But over there, I notice a lot of different people participating in discussions, a lot of constructive discussions as well. And really, if the topic is taken up, it's progressed really quickly until its conclusion and then makes quite a stake up. So it seems to work really efficiently and really a nice cooperation between the different people participating. So I'd like to move on to sort of, here we have been working at it at this project for about a year. We've done three releases so far. And of course, we've had extraordinary conditions to add a monkey ranch to everything in the middle of it. So we've had to move to be 100% virtual as opposed to our, I think we were very lucky, our last face-to-face workshop in Prague in January of 2020. I think it's totally amazing that we emerged unscathed in retrospect, of course. But moving forward, this was a project that kind of came out of the woodwork. And absolutely the Linux Networking Foundation and GSMA have been very supportive. But it has been kind of a skunkworks. And so it really needs to have a place to go. So coming as we move into our second year, we're actually moving forward and taking some next steps. And I would like to open it up to Scott and Tom and I were going to be talking about sort of what happened and where we're going with the project. So right from the beginning, there was always talk that the CNTT project would be using the OPNFV group. Because OPNFV has been involved in creating labs and testing environments and implementations. So it was a logical fit. But there was also a thinking that the OPNFV group tended to be more reactive in the sense that it didn't do a whole lot of architectures and reference models. So there was some activity around that. But obviously there was enough of a gap that CNTT became a full blown project very quickly. So I want to talk a little bit about the decision-making principles behind where we're going with this. We really needed to make a decision and move forward. And Scott, I know you were involved in this. If you'd like to, and Tom, both of you would like to comment on this before we go into the actual decision. Yeah, sure. Any time you bring lots of people together, there is bound to be difference of opinion. And how should we do this or how should we do that? And this is no different. And when you take established communities, bring them together, and talk about, well, what do we want this to become going forward? To do that, we had to start with an agreement. This is what we want to get out of it. So to facilitate that, the group joined up together these principles. And it's important because both groups need to continue on doing what they're doing, but doing it together to make a greater whole work better. The phrase, better together, I think, applies here. So I think we've hit a lot of the points here. One that I like to call out is the third point there, ensure a continuity of work. Coming together in two groups into one, it makes a lot of sense when we think about architecture requirements, engineering, building it, testing it, typical type of life cycle. And I really think that third point, let's not disrupt. Let's keep the work going on the community because in the end, that's what stands. So we've got to keep that work going. And at the same time, let's rebuild into a new group that can do better together than we were doing individually as separate. So let me move on to, so there was a group that was formed that represented OPNFV, GSMA, LF Networking Board participation, as well as CNTT. So it was a joint group that came together to come up with how we should move forward. And it was sort of interesting with dynamic because for a long time, the CNTT people were kind of reluctant to work particularly around with the idea that the OPNFV and CNTT really needed to merge. And so we went through the exercise. We looked at keeping CNTT as a committee, didn't make a whole lot of sense as a project, also didn't make a whole lot of sense. There was a lot of heavy lifting that the expertise of CNTT didn't have. So there was an epiphany where we realized that it made sense for OPNFV and CNTT to be really reimagined as a new entity. And that's the exciting work that we're going to be working on over the next few months. And Tom, would you like to comment on that? Because I know you're participating in this very heavily. Yeah, I think the work to actually discuss how we will do this merger really needs to start in depth now. So on that, there's not a whole lot of detail yet. But it's clear that that's the path we want to go on to. We've also received confirmation from other boards that they're supportive of this approach. That's really where we want to go to. But as you said, Scott, the intention is not to disrupt existing activities. And that work should continue. So this will all happen in parallel to what we're doing. And we will try to define a new organization that this existing work can flow into nicely without accepting it too much. And then start working together more in depth. This will be alongside what we're also still doing with the GSMA. So the reference model, as touched upon earlier, is going to be published as a GSMA specification in order to guarantee stability of that document. That will continue even in this new setup. We've established a group on the GSMA site. As soon as we do that, this will continue also in the search because of the transition of the merger of European fee and CNCT. And I'd like to also add that I know GSMA has just kicked off some edge activities as well. And the CNCT has an edge work stream. And I know that there's already talks about those two activities working together as well. So it's definitely better together as we have. So, and Mark, I'd like to have you talk a little bit about, before we go into the sort of details of the transition roadmap, I'd like you to talk a little bit about the activities around the certification a little bit more because that's probably the least fleshed out part of the project so far. And we're hoping that with the merger that this will give impetus to make that component become more prominent. Yes, I think certification is going to play a pretty significant role when we take that step toward operationalization of CNCT. But even before so, the fact that we're going to be rigorously verifying or at least creating the functional requirements are going to enable us an opportunity to validate even the specifications for CNCT on the front end. And then as we become more familiar over time, the certification and testing program will allow for operators as well as vendors to be able to differentiate their offerings by providing CNCT conformance. And at the same time, it will help take that major leap forward toward multi-vendor CNCT, which is the ultimate goal of any open community. It's not just about one entity creating that conformance. It's about the overall community collaborating and being able to build with multi-vendor building blocks, which is what CNHouse certification can actually emerge. Now at this stage, though, it's a very early in that discussion. So there are a lot of resources that we have to leverage. The OPNFV community provides a lot of extensive testing, but they also provide integrated scenarios for these predefined stacks of Linux Foundation open source networking projects that we can use in not just the definition, but in the conformance. And those integrations are not trivial to achieve, so we definitely need to think about how we can leverage that, those integrations, as well as the testing projects in the OPNFV platform. But then, of course, we're going to have to look and tailor the certification activities around the specific needs, and we'll have to hold off on that a little bit until things solidify with the new merged project. And so Scott, if you could talk a little bit about this roadmap, and then we'll talk about how you can get involved and what it takes. Yep, so we're here at the end of August by the time this will be viewed, it'll be September already. And really, our main goal to take away from this one is that we'd like to have this up and running by January of 2021. So to do that, you can see throughout this, there's a lot of steps to get this to work out together. Now that we've come to the agreement, yes, coming together as a single group is really important. And that ranges from anything from figuring out the organizational structure, the new organizational structure, to what are the names going to be, press releases, just electing the right leadership, and actually getting a public launch ready. So there's a lot to be done, but the teams are committed to work together and have this ready in January during the first quarter of 2021. And I should point out that when this is being shown in September, there will be opportunity for anybody who's listening to this to join the project. So feel free. And with that, I'm going to transition to the last slide where we're going to talk about how you can get involved and where the materials and information about the project are kept. So I am opening it up to the group to talk about what's the best way you think that other people can or other organizations can join. And Gergay, and you're coming from the vendor perspective, could you talk up for a minute or two about that? Yes, so I think for vendors, it's important that this is an open source initiative. So we can freely comment and modify the specifications. And it's very easy to create issues on GitHub. So if there is any problem with the specifications, anybody is fixing the issues. This makes that it's a leveled activity between operators and vendors. So we are equal partners in this work. That's very important. And Scott, would you like to comment on that from the telco perspective? Yes, obviously the operators will be deploying all the infrastructure. So there is a huge financial burden there to carry to make sure standards are right. But the ecosystem is not just the operators. It's got to be everybody involved from any supplier, whether it be a VNF vendor, a CNF vendor, test suites. And there's so much just in between that has to make this right. And really at the end of the day, we're doing this to make it easier for everybody to come together, agree on an approach. Because this is not the area that differentiates one company or another. There's enough tooling. There's enough standards to draw upon, bring it all together in one place. So what you see on the screen here is a lot of places to go and get together, to come together to contribute, even if just starting at a weekly governance meeting to come in. You can start by going to the CNCT home and navigate from there. You can see the meeting structure and come to a governance call, hear what's going on, get a feel for who's involved, how it comes together. There's a technical meeting that happens. We come together on Wednesdays just before the governance. And then there's a bunch of technical calls throughout that hit all the one stream, the two stream, and the edge topics. And those are scattered throughout the week. I also remind you that there's a whole onboarding process available as well. And so that can also be found in the quick reference guide. So there's a whole process for getting new users to come on board and get up to speed quickly. So that's very important. And we are coming to the end of the meeting. At the session, thank you everybody for participating and hopefully you learned a little bit more about how we are creating reference models and reference architectures for infrastructure to support the NFEI, or sorry, the NFEES, the network function virtualization efforts that are part of the LFN and the other open source community as well. With that, we are moving forward.