 and JSA Radio, Your Voice for Tech and Telecom on iHeartRadio. I'm Jamie Scott-Okitaya, and on behalf of my team here at JSA, welcome to our monthly virtual roundtable. We are bringing together top thought leaders, talking about topics important to our industry in our monthly virtual roundtable series, available right here on our JSA TV YouTube channel, as well as on JSA Radio, the only tech and telecom podcast series currently available on iHeartRadio. These monthly roundtables lead us up to our on-site CEO roundtables at our executive networking event, the Telecom Exchange. Next one's coming up fast, June 20th is the 21st. That's at Cipriani's Wall Street, downtown New York City. More info at thetelecomexchange.com. Today's virtual roundtable is brought to you by our video collaboration managed service provider, Pinnicka. With your video platform, our panelists are able to stream live from all across the world. So thank you, Pinnicka, and thank you to our viewers who are joining us live and to those joining us on demand. So let's go ahead and get started. Today's topic, SDN and NFV, myths and realities, helping us to break this down. Very honored to introduce to you our guest moderator and my friend, Carol Wilson. She is the editor-at-large of one of our top industry media sources, Light Reading. Carol has spent over 25 years covering telecom, including at Telephony, Interactive Week, the Net Economy, and her own telecom news website that she founded, Broadband Edge. In her current role as editor-at-large at Light Reading, she's the link between the editorial team and other parts of the Light Reading Empire. She's also the dean of the Light Reading University. Her beats, you guessed it, SDN and NFV, making her our perfect moderator for today. Carol, thanks for joining us today and thank you for your kind insight. Thank you, Jamie. These are very kind words. I'm excited about today's event because we have two industry experts who are actually on the front lines of NFV and SDN. We spend a lot of time, I spent years already, writing about how this technology has come about. But these are two individuals who are actually hoping to deliver it into the market and make it work to deliver services. And I'm going to let Nyla Far and Dean introduce themselves and have each of them tell us a little bit about what they do so that we get a good perspective on the conversation that's to come. So Nyla Far, why don't you kick it off? Thank you, Carol. Hello, everybody. I'm very excited to be part of this virtual round table. As Carol mentioned, I am very involved in delivery of SDN and NFV services to the market. I'm the director of global consulting at CNO. My function involves working with our service provider customers in the business units, mostly product managers, marketing, CFOs, to define how these SDN and NFV services will be described, how they will be priced, the pricing strategies, and what are the business cases that go to market methods and strategies and market opportunities for SDN and NFV. OK, great. Thank you, Dean. Yes, thanks, Carol. I'm the CTO for Light River Technologies. We're a network integrator that works with a variety of customers across several market places. And we're at the front lines. We're actually trying to help some of the tier two, the mid-sized and smaller guys figure out what's real for them, what can provide value for them. It's clear that the big guys have the resources to really heavily invest and try to make this stuff real and workable. But we're working with the mid-sized and small guys trying to figure out how to make it workable for them and how to drive it into the mainstream of the marketplace. Great, thank you. So I want to kick this off by just asking you both sort of a general question about the pace at which NFV and SDN are being rolled out. I've been writing about this stuff for years. You guys have been heavily involved with it also during that time period. And there's some skeptics around it would say this isn't going fast enough or telecom needs to do it differently. What's your perspective as folks, as you both said, are kind of out there on the front lines of delivering this? Is the transition going as you expected, as fast as you expected, or are there things that you think could speed it up? And none of our ones should go first. All right, very good question. So let's understand that I think we all agree that the telecom environment is fairly complex. If you look at the history of major transformations, take, for example, the move from voice centric to G and 3G to 4G mobile broadband, or the move from DSL to expanding fiber deeper with the fiber to the premise. These major transformations took several years to be realized. So I think SDN and NFV is no exception. It will take some time. And the reason is that the network and operations of the telecom industry is fairly complex. Now, I want to bring, rather than focusing on the bigger picture, I want to bring some tangible examples. So let's look at the status of the industry. We have some providers, like CenturyLink, that have been very vocal about their plans. They announced their programmable services back on a couple of years ago. And as you know, they have been progressing very well, indicating that they are well ahead. And they have indicated 40% of their core IP has already become virtualized and so on. Or we have orange business services where they have started to offer a pilot for virtualized enterprise services in 2015. And in 2016, they have made it available as a market launch under the brand of EZGo. So there are many examples that some providers, large tier one carriers, that are pushing forward and announcing their plans. Now, I want to take a moment to look at what are the motivators for the providers and also what do we see as challenges and what is really the slowing factor. Motivation, as you know, is that with the convergence of the cloud and network, operators realize that they need to automate. They need to bring services that have more connection to the cloud and create a network that will be orchestrated and operated similar to the cloud environment for the enterprise and consumers. So the business incentives are there. And I'll take maybe later on, I will elaborate more on the business aspects. But the challenges are related to integration aspects. Even though open source environment creates an environment which is promoting collaboration, there continues to be certain level of integration with the operating support systems and the building support systems. These systems are fairly wide spaced. They have a combination of in-house developed as well as third party developed software. So it would not be realistic to think that over time these systems can integrate with the southbound interfaces that are defined in SDN and affiliate here. They need to stick some integration effort on the northbound interfaces to make this a fully functional, business functional solution. And that's one aspect that is elaborating on the challenges that we see. There's also the way these services need to be priced. The business models for these services continues to be a question for the carriers. And they are defining those frameworks. And lastly, suppliers also are somewhat in a similar situation because most of the suppliers who are providing the virtual network functions, they have been traditionally in the hardware business. So the industry has to learn to work together to make this progression. But the good news, I think if you look at it half full part of the glass, the good news is that it is progressing. And we see operators that continue to make their announcements and their intentions widely, publicly available. Great. So you've really hit a lot of the key issues. And I want to come back to you on some of those. So I'm going to ask Dean, from your standpoint, dealing with the smaller operators, the tier 2s and smaller, what are some of the challenges they're facing? What are some of the things you think they might be able to do to speed up their own adoption of SDN and NFE? Yeah, Carol, I think Nilafar actually hit on a couple of the main topics. Are things progressing as fast as we expect them, yes? And is it fast enough? And the answer is no, always. And I think part of that is it's a little disappointing to me that even after the last couple of years of relatively intense work and focus, a lot of the marketplace is still really talking about technology and not the actual solutions. So we see a lot of companies out there focused on re-architecting around microservices and using Docker and really focusing on kind of lower levels of the stack instead of getting to the point where Nilafar was talking about actual integration with the other operational systems. To some extent, I think a lot of this technology coming out of the data center world, where the data center challenge was about scale rather than managing 10 switches, managing 10,000, when you come to the telecommunication space, it's about managing complexity rather than just scale. And I think that's what's really hit the marketplace pretty quickly was they thought they could just do everything they do in the data center and run it out to the telecom operators. And they found out that their environments are significantly more complex. So everybody wants it to be moving faster, but it's been a challenge to try to get the technology to the place where we're actually talking about return on investment and how to actually implement it in the environment rather than putting together, doing all the integration work, the low-level work, just to get the baseline applications in place. OK, well, I'm going to follow up with both of you on that last point you made, because it ties into something that Neelafar said about business models. What do you think, or what are you seeing as folks that are out there trying to make this work in terms of business drivers and business services that can be delivered today potentially in advance of solving all these background integration issues? Because, Dean, as you pointed out, the complexity of telecom networks is amazing and widespread. And if you wait for all those issues to be solved, chances are you're going to go out of business in the process. So what are some of the things that can be delivered today, and what are some of the business models you see that are actually working? So what we see is a great deal of focus right now on NFV, and particularly at the Edge. So a lot of the service providers are looking at being able to deploy CPE to their customer site that is much more flexible and gives them a number of different ways to monetize their network. So this is the use cases everybody always talks about. Being able to one day have just a customer in the next day drop a firewall onto their CPE programmatically and suddenly be offering a firewall service, a managed firewall service, or encryption, or WAN acceleration. We see a lot of focus, particularly around the regional, mid-sized, and smaller companies, focusing on that because it's very, very fast return on investment. Interestingly, one of the things we're running into right now is it's actually being looked at as a way to deploy SD-WAN applications is with the virtual CPE. So we're seeing kind of the convergence of those two big waves hitting a lot of the mid-sized and small guys. The jury's still a little bit out on what we call carrier SD-WAN, which is the overall network automation. We're still seeing a lot of slow going there. A lot of guys trying to get, when they say multivendor, trying to get that number over two is always a challenge when most of our telecom networks have tens of vendors in them and not ones and twos. So. Question about how is the SD-WAN market done because everybody's picked their vendor? And I'm like, when is telecom ever gone with a one-vendor approach? But that seems to be where we're starting with SD-WAN. So Neelafar, what's your perspective? I think Dean's raised that definitely what we're seeing is virtual CPE because there's a strong business case for that, right, in terms of reducing the number of boxes on the crime, getting services to enterprises faster. What are some of the other things you're seeing? I absolutely concur with Dean's observation. We started to look at the VCPE map a few years ago. And today, I'm going to candidly tell you that we receive requests for information, requests for codes, requests for prices from the telecom operators. Which indicates that things are getting fairly realistic in terms of rollouts of the virtual CPE. So the business case for virtual CPE is quite clear. But now let's understand once it's become a virtual environment, reduction of the boxes and reduction in the operations aspects of it and truck rolls and the reduction of cost. What would be the real value of offering this virtual CPE is to create an enterprise marketplace. If at the end of the day, we end up having just one single routing function or one single firewall function or one single SDVAN function on that virtual CPE, it kind of defeats the purpose of going with virtualization. We have reduced the cost. And that's a check mark on that. But have we created the best of breed or have we created options, the ultimate option for the user to select what they need, dependent on they are connecting to a cloud environment. They are out of their geography. So these are factors that are important as a strategy for the virtual CPEs rolling out. Is there a marketplace? Is there a multi-vendor strategy in place to enable the provider to ultimately differentiate their VCP, not to reduce it just as yet another box, but rather a device that connects the users to a marketplace where they can get the best of breed applications? That's an excellent point. What's it going to take to get from where we are now to where you have that marketplace, where you can, as a telecom operator, actually give your customer a choice of best of breed options that I assume would be delivered on demand from a cloud? That's it. OK, so what's it going to take to get there? So what does it take? I guess you know the answer from a technology perspective. That's the end and if you are complementary, we need to create APIs, interfaces for customer portals, self-service portals. Although we have talked about the service, self-service portals for a long time, I'm not quite sure if everybody has rolled out self-service portals for their enterprise customers. So one aspect is to make it easier for the users to select these functions, which goes back to the platforms for SDN and NFV orchestration. Create an open environment where resource adapters can talk to multiple vendors and allow these functions to be provisioned into it. But before we even go that far, I want to bring us back into realities of today. Let's take a look at what is the market offers for enterprise connectivity services. We are looking at a good growth of Ethernet services, layer 2 services delivered at Ethernet, at the Ethernet CPEs or needs. So these services are growing according to many analysts at double-digit factors, 12% or more. But have telecom operators truly automated the provisioning of the connectivity before we even get to the VCP? I think we have to do some first steps before we get to the VCP. And I call that van automation. Have we automated the fulfillment process of connectivity services? Have we improved the services assurance? When a trouble ticket comes, do we look at multiple panes of glass to figure out what went wrong? Or do we look at single pane of glass? These are very immediate steps that need to be taken before we even get to talking about marketplace and cloud-based network services. Let's automate the provisioning. Let's automate and improve the service assurance. And then we can talk about those value-added services. OK, so, Dean, would you agree with that? And are you seeing any carriers out there today actually tackling that challenge? So it's interesting. We're talking about, again, kind of the multi-vendor being able to do provisioning at a high level with either automated provisioning, customer-driven provisioning. And it's interesting to me, the idea that we're going to get to a model where we can actually do multi-vendor provisioning across the complex telecom networks very quickly is still one of the things that eludes me. We did it with switches in the data center. And to this day, we're seeing that kind of infrastructure being available where it's pushed a click and you get services instantly turned up. I hear a number of my customers every day talk about what Amazon can do with AWS literally in 20 seconds is where they want to get to. But the big guys, the content providers, the Googles of the world, and the Facebooks had enough market power to drive the switching vendors and create their own switching platform to do that. I have yet to see any kind of standardization across routers, optical devices, any one of the other 10 or 20 different network elements that would be in a telecom provider's network. So that challenge to try to build multiple communication mechanisms to drive it through a single provisioning system still eludes me about how we're going to get there in the near term without just a lot of brute force effort. But I do agree. We're seeing a lot of customers really trying to focus on preparing their networks to get there. They're looking at the different orchestrators that are out there and trying to pick maybe not orchestrating and automating their entire network, but the deployment of one service or the deployment of one DNF or two. The being able to automate those, again, with a minimal investment gives them some return on investment that's viable for their business cases. So yeah, I agree. We're all driving that way. We're all driving toward the network automation space. But I'm a little bit skeptical that we're going to get there anywhere in the near term. We see customers doing the NFV, particularly the VCPE type of stuff, beginning to do those without a lot of automation, because it gives them, quite frankly, value from day one being able to lessen truck rolls and some of the stuff, some of the physical costs around their access networks right now. That makes sense. So I think this is a good point in the conversation to bring up the notion of open source and the role that open source is playing, because we start talking about automation. We start talking about multi-vendor environments. We are seeing telecom embrace, open source more. We're seeing the growth of multiple open source MANO groups, including the open network automation platform and the open source MANO. So I'd like to get each of you. And Dean, why don't we start with you? Your perspective on the role those are going to play in enabling some of this automation, enabling some of this multi-vendor environment. Yeah, so I've been actually pretty pleased with what's happened over about the last year with the increase in focus on some of the standards groups. I'm always a bit skeptical that any vendor is going to provide the open interfaces that we need, because that eliminates their ability to differentiate themselves in the marketplaces, right? But I've been actually pretty happy with the increase in focus on standards, on trying to get some of these common languages defined so that we can talk to multiple vendors without writing custom code for every vendor. Speaking about open source directly, it's kind of interesting as we've watched some of the development from the different vendors actually on the SD-WAN, carrier SDN front, and even the NFV front, we've watched them sort of work through the stack as they started at the lower levels, realized that an SDN controller doesn't provide a lot of intrinsic value or differentiation. So they moved their development efforts away from the controller wars that were two or three years ago to suddenly everybody wanted to be an orchestrator. And now they've realized that an orchestrator, even in many cases, is kind of standard APIs in one end, standard out the other end. So now people are beginning to focus on the higher levels, the orchestrators, and the domain level, a higher domain level, multi-domain level orchestrators, and trying to get into that workflow process, that overall, almost the OSS-VSS space. So we watched as people have tried to find a spot where they can actually add differentiation, because open source seems to take up a good deal of those niches where they'd like to develop their own custom code. Now, the one thing I will say about open source that's concerned me is the support model. Right now, in developing some of these applications, when we sit down and talk with a customer, we're actually forced to tell them, hey, if we develop this application the way that the software is constructed right now, your support contract is going to be with four or five or six different entities in order to try to get technical support on your application. That's not a tenable support model for the long term. So we've got to begin to look at some way of how do we, going down kind of the red hat, Linux type of model, how do we support a lot of these open source components? How do we bundle support for an overall SD-WAN, S-Carrier SDN, or NFV application? How do we bundle support through one person or one entity rather than dealing with the vendor that created it, the three or four open source companies that are part of their stack, the individual elements at the bottom of it. So open source playing an incredibly valuable part, particularly around standards, and as I think it's only going to get bigger, but we've got to work out the support model and how it's going to be supported over the long term in a commercial environment. So, Nilefa, I think it's an interesting question to bring to you, because Santa has in at least one case I know of, delivered a hardened version of an open source platform and maybe you could address the things Deans raised in terms of the support models going forward for open source. So we are a big supporter of open source and I think you know that, you have tracked what we do in Blue Planet is we absolutely believe in open source, we believe in community. As soon as the code becomes available, we take it, we augment it, we make it available to the community. And I think it's an important piece of desegregation of the hard way and software, that's the philosophy behind it. Now, we do have a strategy set solutions like DevOps tool case. So the DevOps would enable the best practices in software development to be combined with the best practices in network operations. And we have different models for availability of those support models. For example, DevOps can be different levels of training available to our carrier customers. I concur within that the support models would vary depending on what the carrier needs are. We have seen a range, those who want to be trained on the DevOps and then would like to have as flexibility as much as possible. And it has been one of the strengths of the platform to be less integrated and more flexible and allow the development teams to make changes fast enough. This was one of the key factors in selection of our platform in some of the large deployments, the flexibility and not being too tightly integrated and allowing the development teams to try things and develop their own. There's also, we also see need for cases where a higher degree of integration is required. So it would range depending on a case by case basis and what the carrier is really desiring from a support perspective. We have different tiers for that. Okay, all right, that makes sense. But I think fundamentally with open source, you have to have somebody who's tracking things like version controls and patches and keeping code up to date and things like that. I mean, I'm assuming, and this is for either one of you, that your carrier customers are looking to you to help them do that as opposed to doing it themselves. So that's one of the challenges in the marketplace right now is we're getting in and starting to talk about actually building applications, provisioning and automation applications for customers. As an integrator, I mean, it's part of our role, but as an interpreter, yes, we're looking at either passing through or supporting four, five, six different components within a single application for a vendor, for a customer. It's an interesting development in the marketplace. It's challenging us. We're attempting to rise to the challenge and we're looking at, as Neil LaFar was talking about, different models depending on the customer, how much control they want and how much just pure integration they want. But it's clear that at the end of the day, almost all of them, particularly at the executive level that I talked to really don't want to become software developers. They want to focus on their own business. So the flexibility and they want the flexibility, of course, they want the agility to modify and offer new services, but they certainly don't want to become a large software development house as their primary business. Seeing challenge in the marketplace right now, it's not been solved. Okay. All right, well, this half hour conversation has gone by very fast as I kind of knew it would. So I'm going to give you each, I'm going to do a final wrap up question and ask you to keep your answers fairly brief here. And it's an easy question. What are one or two things you expect to see this year that we might note as pushing NFE and SDN deployment forward? So that's an easy one, right? Who wants to go first? I can go first if that's the case. Okay, so these are the things that I expect to really accelerate this year. What I refer to as man automation, which means that a streamlining, improving and automating the services fulfillment. When service order is submitted, it triggers many processes and we already have deployments of our blue planet in what we call multi-domain service orchestration, MDSO for man automation. I think that trend is going to continue to be a focus in the near term because it's quite essential to bring SDN and NFE. Virtual CPEs definitely is getting a lot of momentum this year and I think we would not be surprised to see even higher deployments of the CPEs across the telecom operators. And with the combination of the two, I think we are now moving into emergence of technologies that relate to analytics because down the road, once things have become software defined and virtualized, the truth and the reality is that it's not going to be a human job to configure, manage, change these things and maybe not even a simple software. So analytics is gaining a lot of attention right now and I think we are going to see a lot of exciting things from this. So by automation, VCP and analytics would be my top three bits. So Dean, she gave us three. You've got three ones that she has already chosen. So I think or at least I'm hoping that this is the year that the software stack, the underlying components, most of that is basically shaking itself out with all the discussion around Docker, microservices and all the stuff that at the end of the day is interesting to the software programmers but not really to the guys running the business, right? But that's basically shaking out and this is the year that we're actually going to see some real environments spring up that actually provide the business functionality. Particularly around NFV, we'd be really thrilled and we're kind of watching a couple people closely with their ability to roll out a true NFV managed environment where you can do your NFV or your VMs in the cloud, the private cloud or down on your customer's CPE and really get that flexibility in the access part of the network. We're very excited to see some of the releases in software that I think that'll be coming out this year in that space in particular. And we're going to see the first pass, I believe in what Neil LeFar talked about is that the SD or carrier SDN network automation. We're seeing a number of customers with roadmaps that show their initial release, again, multi-vendor maybe two but at least it is multi-vendor and it is a framework that can be extended or their software is framework-based so that it can be extended to begin to deal with some of the complexity in the carrier market today. So I'm really hopeful this year is going to be the year that we really begin to see what I tend to call the kind of the out of the box applications that can really provide value. The other thing that we haven't talked too much about and the industry really hasn't talked too much about because we're really at the beginning phases of it is the ability to get applications environments set up that are truly mission-critical. To date, we've seen a lot of applications out there that are great until they crash or that work well except they don't have those monitoring Neil LeFar mentioned analytics is kind of one lens to look at that through but being able to get applications that are truly fully redundant, carrier class, mission-critical, all the buzzwords that you would expect particularly in being able to manage a virtual environment that way that we manage the physical environment today. I'm very hopeful and very eager to see some of the software developers come out with software that's more than just sort of runs and now we can say it runs in a highly available redundant type of environment. Those are, that's one of the critical things that I see. Before anybody's gonna bet 30 or 40% of their revenue on it, they want an environment that's not going to crash, that's going to be able to be fully redundant as redundant as their hardware networks they built today are. Redefinition of what Five Nines is in the virtual world. Exactly, yep, exactly. Great, well, I wanna thank you, Neil LeFar and Dean, for your participation today, some great insight. I think I'm gonna hand this back to Jamie at this point. Thank you, Carol. And thank you, Dean and Neil LeFar, what a helpful and insightful conversation. Really appreciate your time and thoughts there. We hope to see you live at Telecom Exchange in New York City, June 20th through the 21st. And thank you, audience, for joining us. Check out more of our virtual roundtables on jamescato.com and our live CEO roundtable lineup at the telecomexchange.com. If you'd like your C-level to be featured here next time, please email us at pr at jamescato.com. Thanks everyone for tuning in to JSA TV. Be newsroom for tech and telecom professionals and JSA radio, your voice for tech and telecom on iHeart radio. Until next time, happy networking.