 Live from Silicon Valley, it's theCUBE. Covering Mobile World Congress 2017, brought to you by Intel. Okay, welcome back everyone. We're here live in Palo Alto for theCUBE's special broadcast presentation and coverage of Mobile World Congress, which is happening in Barcelona, Spain. I'm John Furrier here with SiliconANGLE Media is theCUBE. And of course, we're covering it here in Palo Alto, bringing in experts and friends who are following all the action, as well as have commentary and opinion on what's happening. We're going to roll up the news. It's the end of the day in Barcelona. We're just getting our sea legs here for day two of 8 a.m. to 6 p.m. coverage inside theCUBE. And of course, we want to break down the counter. Our next guest is Val Berticovici, who is the CTO at SolidFire. Also a governing board member of the CNCA, the Cloud Native Compute Foundation, which was KubeCon, which is now part of the Linux Foundation, which if you know theCUBE, you know we've been covering that like a blanket. All this shows theCUBE has been there. This is in the world of DockerCon, et cetera, et cetera. Val, a CTO, 19 year veteran at NetApp, of course, knows the storage business, knows the converging infrastructure, knows the cloud, part of the original cloud already. Val, great to see you. Thanks for coming in. Thanks for having me back on after all these years. Yeah, I mean, it's great to have you on. One, you know, we've, you know, we see each other at some of the parties and Georgiana's place in particular. Georgiana brings all the storage and cloud already together. But you and I have had conversations a few years ago about, you know, where cloud was going and you almost kind of connect the dots, not to pat ourselves in the back, but I think we were right that the cloud was what we thought it would be and probably more. I mean, for me, I think I underestimated certainly the Amazon impact. Me too. But you look at what's happening in Mobile World Congress, you have a bipolar show. You have a device show, ra ra, look at the fancy devices. And the other show, you have a telco show that's trying to figure out their future. And that's interesting because the telco's power, the big networks that everyone is using the devices for. So you have a consumer market, but the real conversation is 5G, IOT, you have a collision course of enterprise issues, enterprise data center, enterprise technology, colliding with a telco infrastructure, AKA mobile, head on. So it's not just more wireless, 5G, certainly the story we talked to folks like Intel and others around that, but you have essentially all these core problems that are gonna scale up this next generation use cases are enterprise like. This is your wheelhouse. So are you like looking at this saying, I've seen this movie before, what's your thoughts? Actually, I haven't, that's so exciting, right? There's, as you said, there's so much innovation happening for me. Probably the big story is what's not in the headlines at Mobile World Congress, which is the back office work to support a 5G rollout. And I've had a lot of experience, particularly on a solid fireside most recently, speaking with all the big telcos globally, and their NFV implementations. And what's interesting is they're all going through a gen two or a rearchitecture right now. A lot of first generation NFV was done based on traditional or legacy now cloud architecture, which is very VM based. And all of them are now architecting and implementing microservices based implementations. And a driver for that is just the explosion that 5G will enable in terms of connectivity between devices. So the least interesting stat to me is how fast I can download a movie off of 5G. The most interesting is how many hundreds of thousands or millions of devices within my domain are going to be communicating with each other on 5G. We had Sark Ghalai, who's going to come back today. He's also a guest analyst, former HPE, NHP Enterprise, he ran there, built their communications group all the NFV. I like it, and he was commenting the same thing, and he made a point I want to bring up, which is I don't really need more, a gigabit right now, I want more battery life. So he's kind of being pedestrian in his use case, but that really is kind of the consumer issue. You're pointing about things that are going to be harder to do. And NFV, you mentioned as one of them, can you explain the NFV current situation? Because also we've been doing a lot of the OpenStack, all the OpenStack shows since it started, that has become kind of a telco NFV storage show as well. Absolutely. So what is the real issue with NFV, and why is it important and relevant to the service providers right now? So if you take a look at all the services we depend on our phones nowadays, there's obviously the basic connectivity, there's additional services around location mapping for GPS, the layered service is not top of that in terms of the collaborative apps that we use and depend on every day, sometimes on S3, which is not always available as we're recording right now. There's a lot of layers there, and from an NFV perspective, from a backend data center perspective, everything amounts to a session. So even though it's packet switched, it's still a logical session you have to set up. So for every session, and imagine this happening millions of times at every tower and more than millions of times at every regional or central data center, you've got to have a session set up where you've got to authenticate, actually, who the user or the device is, make sure they have permission to be on the network and accessing certain things, you've got to authorize them to do certain things, you've got to log what's happening, then you've got to slap some firewall or security around them, then you've got to layer in and access to all the other resources, they're trying to combine into a service back to the end user. There's a lot of things going on. You have to set up these sessions for every connection. And if you try to set up a VM for every connection, you'd have to fund a multi-billion dollar data center, Google can't even afford. So this is where microservices are becoming essential right now in a 5G hyper-connected world, is where you have to have much more efficiency in the speed by which you set up these sessions, the efficiency of number of sessions per server, and just the end-to-end cost of processing all these transactions. This is interesting, and I want to just kind of translate a little bit out for the folks that aren't CTOs out there, essentially, you think about mobile. We've all been, you know, since the iPhone in 2007, we've seen this just accelerate with data and whatnot. You've been at a concert, you've been at a stadium, and you can't get, you got signal, but you can't connect. That is essentially the base station saying I can't get a session. Now, as a user, you have a phone, so you've been provisioned by the telco. You actually, they know who you are, so you have a phone, you have a device, you just can't connect a session to the radio connector, and then get to the internet. That's a known problem. Now, when you think about IoT, Internet of Things, and now people, your watches, your wearables, sensors on airplanes, and industrial equipment, to traffic lights, those are devices that are going to be provisioned and turned off and on, so it's like a new phone every time, so you've got the complication of not knowing the devices that are coming on, and then trying to establish the connections, and figuring all this out, this is kind of a really hard problem. It is. At scale, it's really, really hard. This is a really hard problem at scale, at many levels. So to me, what we're hearing at Mobile World Congress is, you need a dynamic network. Absolutely. What are some of the tech involved? What's the real enables? You mentioned microservices, we know about containers, Linux foundations opening up their kernel for a lot of variety of new kind of configurations. You got solid state memory, and you got new memory architectures. What are some of the key things from a technical perspective that are going to change that complexity to be seamless for users? Probably the most fascinating trend to me, and we're just beginning to see some stories and a merger on this is the rise of edge computing. So I kind of hark back to where I started my career and dating myself now, but the client server era that succeeded from mainframes, we've seen a huge pendulum shift towards cloud computing and centralizing a lot of processing. Well, back to 5G, back to millions, if not billions of connected devices right now, there is no way Einstein kind of introduces problem for a speed of light. There isn't a way to process the exponential amount of data we're dealing with right now at the core, and still provide useful feedback at the edge. So the rise of edge computing as a bit of a counterbalance to cloud computing and having more powerful, more intelligent processing at the edge, filtering a lot of data because we can't possibly store the exabytes and the other bytes of data. This is a paradigm shift. What you're talking about is a new paradigm shift because it used to be a centralized computer and a master slave or connected device terminal, then you had smart terminals, which is clients and then PCs and then you got smart phones. So what you're bringing up is an interesting architecture that is an enterprise data center thing. We were talking yesterday and then I was talking to the Intel folks, I pressed them on this because they're obviously in the data center business, that a car that's fully instrumented like a Tesla or a future autonomous car is essentially a moving data center. So it just changes the notion of data. Yes. This is a paradigm shift. You agree with that? Moreover, IOT is maybe the first technology buzzword that takes a lot of this digital world that we've been talking about. It's really been largely abstract and virtual for the common person, and it makes it physical and real. So the impact of IOT is an actuator changing your traffic light. You know, it's whether you're getting water or you're getting electricity at your house, whether you're finding your way to, you know, to a new location via GPS. That's actually impacting your physical world. It's no longer just a virtual thing. So that's where IOT is going to become really, really significant in our lives. And the software program that needs to be created, this is an opportunity for entrepreneurs. Certainly, you know, Peter Barris and I were talking yesterday morning about the edge computing and he's got a big slew of research on this. And he talks about IOT and P, IOT and things and people. And we were also talking yesterday about the relationship of the people to technology. So for instance, in telcos, they view the phone as the relationship that's coupled to the carrier. And that, you know, the premise we put out yesterday was that that's going to be an uncoupling. It's going to be a person's relationship to multiple carriers, if you will. So the question is, how does business extract the value out of this? And this is something that, again, Peter Barris and Ricky Wands are digging into, which is the business value of technology. In the paradigm we just talked about, which seems pretty obvious how we get there, not so obvious, but people are working on that. How should companies think about getting value out of this massive shift? A lot of moving parts. The examples are already there, you know. Let's talk about one of the most talked about companies around here, Uber, right? And not for some of the headline reasons they're attracting right now, but some of the classical disruption they enabled. When they take a look at the fundamentals of what they do technically, it's interesting. It's somewhat impressive, but it's not revolutionary. What made them revolutionary is digitizing the transportation relationship you and I have with our transportation providers. And when they tapped into that, they realized the potential of that is limitless right now, because we're all physical beings. We still have to move ourselves or our food or our packages around. But digitizing transportation is really, you know, a great example of any industry, whether it's a 100-year-old industry or brand new industry, when you digitize a distribution and when you actually add digital efficiencies on the back end, you end up with that 100x effect, more than a 10x effect, and truly earn the term disruptor. Give us some more examples that you've seen. I know you mentioned you talked a lot of telcos, but what other use cases? Because the whole notion of 5G and this new architecture is really coming down to use cases. And certainly there's the sexy ones. The car, the smart cities. I mean, there's a lot of policy and societal impact issues that need to be thought through, but just generally, what's the low-hanging fruit right now? You know, instead of low-hanging fruit, I might be giving you the most pedestrian example I could think of, which is when I meet with some of the waste management companies, years ago, I took them for granted. There's no innovation here. It's going to be an old enterprise discussion with the conservative tech leaders. I'm not even going to try and phone this in. I'd show up and meet with them, and they were truly innovating because they realized the whole customer experience of putting out your trash bin, your recycle bin, your organic bin, and so forth, your compost bin can actually be improved and efficiencies added when you put GPS trackers on all the trucks. When you figure out when they have to go to the dump because there's like an exponentially high amount or an inordinately high amount of garbage put out early in the route one morning. And the ability for you to know when your bins are picked up so you can actually go and pick up those bins, put them out minutes before, instead of the night before, bring them back minutes after, just reinventing that very pedestrian-made experience tells me there's opportunities for innovation everywhere in our lives. So really, let's just pick a spot to make efficient. It's probably the easiest, kind of laying back. Great, great feedback, Val. Thoughts on developers, because it's something that we didn't get time, and we'd love to have you bring you back for more time on this. But the CNC, yeah, the CloudNative Compute Foundation is, it kind of came out of the Kubernetes con, which is the Linux, now part of the Linux Foundation. You mentioned microservices, Kubernetes orchestration, there's a lot of software around composability, whether that's an artisan-like developer, or the hardcore developers down low on the stack. How does autonomous vehicles, and this new future use case, whether it's programming drones, or writing cool software, to what's going on in the developer community? Can you share any color on trends around what being done in traditional, classic developers? Autonomous vehicles are really a perfect example because CNC, fundamentally, is about what we call CloudNative technologies and applications, and a typical CloudNative architecture is container package, not VM package, and it's dynamically orchestrated. So we say it's basically declarative, as opposed to opinionated, to use back-end speak, technical developers speak. But what does that mean? Autonomous cars are not interesting if one car is autonomous. Autonomous cars are interesting when dozens, hundreds, perhaps thousands of cars along your route, if those are autonomous, the interaction between them, and making sure they don't all try, and occupy the exact same space at the same time. That's pretty essential. But also keeping traffic flowing smoothly and preventing unnecessary traffic jams, since the coordination of multiple processes, multiple things, that's what really makes CloudNative computing happen at scale. So CNCF actually is five projects right now. Most people know it as the Kubernetes project, and it is very much that. I think a year ago no one knew how hot Kubernetes would be, but certainly it's taken off right now. Thanks a lot, and part to Google. We've got new technologies around monitoring. So you've got to monitor your app, obviously, in the Cloud world to see where the efficiencies and the performances for the end user are on logging, distributed logging, and monitoring. We've got new projects around actually debugging at scale. So debugging one process, as simple as any developer knows, debugging multiple concurrent threads or processes, that's still a black art. And so we've got open tracing technologies around that, and there's a new style of project. It's actually something known as Linkerdee, but it's around now in the IoT context, the most important thing to be a really valuable IoT device is discover your context. What other devices, what other sensors are out there? It sounds like an operating system to me. You got Linkerdee, you got loaders, you got all this orchestration. This is a global operating environment. It's a great, I think, insight. It's not, I don't like operating systems. Not like a classic sense, but there's some systems. It is a new operating environment. I think in the CloudNative world, operating systems still kind of takes one PC or one device or one host on a data center. It really is a coordination of services, new modern high-end services. And declarative things with containers that's essentially assembly-based. You can manage things component-wise, those kinds of concepts. Absolutely. And you see that as a key part of making the CloudNative, as a key part of enabling these new use cases. There's actually no economies of scale if you don't go CloudNative nowadays, as 5G networks become more prevalent, as IoT becomes our mainstream, it doesn't play without microservice architectures. And the trade-off for not being CloudNative is what? Being disrupted. It's literally, you know, there's some great recent blogs. You've heard this title before, the coming SaaSpocalypse. So the disruption of the legacy SaaS vendors. The economics of them forced you to basically have fixed subscription models with your customers. And whether you're using, you know, your CRM app once a day or 100 times a day, it costs you the same. These new CloudNative architectures are going to enable disruption in the industry because they'll only consume resources as a sessions for an app, you know, as common SaaS app are used. And the licensing and business model can now be that much more efficient, 10X more efficient for people that actually want to get charged per use as opposed to per subscription. Val, you know, I've gotten to know you over the years and a great guest to have on theCUBE. Thanks for sharing that insight. But it is a robust, it's exciting time. You had a great run at NetApp. I mean, you look at what NetApp's been doing. I mean, they were the darling of Silicon Valley classic success story. Multiple reinventions, great founding team, great investors, just the classic run that they've had. As you look back now going forward, looking back and now looking forward, what are some of the things that gets you jazzed up right now in terms of things that are, that are that next wave that that's coming? What do you see that's exciting and what would you share for folks for insight? I'd say the most exciting thing to me at a high level is just the opportunity that, you know, 5G IoT enables. I think there's a whole new market segment. Some people might call it the real evolution of HCI, which is edge computing and all these really fascinating new workloads. They're not gonna be necessarily virtual desktops in terms of running or operating your business, but entirely new revenue streams, entirely new services that all sorts of companies, digital or analog can offer. That excites me. And of course we've talked about the backend rise of the shift towards, away from fast storage, fast storage towards persistent memory, that in itself is gonna open up a whole new category of apps that we've yet to see. Yeah, and we gotta get that Linux kernel rewritten and open up all kinds of new stuff. Great, great commentary, Val. Thanks for coming on, Chief Chairman. We certainly want to have you back and really unpack and drill down and double down on what cloud native impact means. And certainly edge and IoT computing really is going to be a fascinating run. I think that's going to open up a huge can of worms and an opportunity for really changing the game and creating great value and risk too. I mean, Amazon S3 is down as we speak. We're joking, but you know, we see insecurity problems out there and we'll stay on top of it. Of course, theCUBE has got you covered and that's the hot themes. Really that's not being reported at Mobile World Congress that we're reporting, which is the surge in IoT relevance. Obviously AI is super hype but that gives a mental model. This is the story of Mobile World Congress, a 5G as a fabric connecting in with hard enterprise data center like technologies end to end for dynamic experiences. This is the challenge for telcos and someone will get it done. Let's see who it will be. Of course we'll be watching it. This is theCUBE with more coverage of Mobile World Congress after the short break.