 Live from Barcelona, Spain, it's theCUBE! Covering Cisco Live 2020. Brought to you by Cisco and its ecosystem partners. Hi buddy, welcome back to theCUBE. The leader in live tech coverage. My name is Dave Vellante with co-host Stu Minim and John Furrier's here. We go out to the events, we extract the signal from the noise and of course it's day one of Cisco Live Barcelona. Very excited to have Prasant Chanois, the vice president of marketing, enterprise networks for IoT and the developer platform at Cisco. Prasant, good to see you. Good to see you folks too. So right now we're in the middle of the DNA center takeover in the DevNet zone. Network's getting more complex. You need a command center to understand what's going on. Give us the update. Why DNA center? Yeah, so this has been a journey for Cisco and for our customers for the last three years or so. A few things happened in the last decade. Like mobile, IoT, cloud and the world of security. All of those came together in one place and if you look at it, these are very network-centric technologies, right? There'd be no cloud without networking or mobile or IoT. So when our customers started investing heavily in the world of applications in the cloud environment, mobile and IoT, the network was slightly left behind. The network that they had created and built was meant for the internet era, not for this multi-cloud, mobile and IoT era. So we had to rethink networking fundamentally from the ground up to how do you help our customers design, build, scale, manage and deploy networks for this new era of digital transformation driven by mobile and cloud. And that was the genesis of our intent-based networking strategy, right? So that was like three years back when we designed a networking architecture that focuses on the business intent and lets you figure out the how part of it, the network figures it out. So the DNA Center was the command center, as Dave, you put it, to help manage, design and build this network from the ground up. And it's been a journey for us and it's been a very, very exciting journey for us where we're getting a lot of positive feedback from the customer, whether it's to deploy their access infrastructure, wired, wireless or moving to the wide area network, extending into data center and public cloud environment. So when we went from internet to the cloud, you know, we talked about the flattening of the network and now I know we're going to talk about it, the edge. Yeah. Are we going to need a new DNA center for that next wave or? No, it's the pendulum swing, right? Like it's always been interesting, mainframe centralized and decentralized edges, then again centralized in the cloud and now cloud moving to the edge. So this is always going to be an interesting phenomena and it's mainly because the world around both sides of the networking has become highly hyper connected and highly dynamic, right? Like users are everywhere, mobile devices are everywhere, applications are everywhere, a single application is split into 500 different pieces, running containers and microservices across four different public clouds and three different data centers, right? Like how do you manage this dynamic environment? How do you set the policy? How do you guarantee an application experience? So this has been a very challenging environment. So the idea of DNA center is to provide you that single command center, right? No matter whether you want to deploy it as a virtual service, a physical service in the cloud, in a hardware platform, doesn't matter, right? So how do you get all of your data? How do you get a single place to provision the system? Well, I'm glad you've mentioned scale quite a few times talking about this. For the longest time it was, how do we get the network people to get off of their CLI and go to the GUI? Well, I don't care if you've got the best GUI in the world. The hyper-connectivity, the amount of changes going on, people can't do this alone. So talk to us a little bit about the tooling, the automation, the APIs, connect all these things and make sure that our people don't become the bottleneck for innovation. Frankly, the complexity has exceeded human scale. It's just impossible. It's funny because I was talking to the CIO of a pretty large global bank, I can't tell the name. Who was saying like, hey, a few years back I had one IT person to manage around 1000 devices, all the devices, right? And then that year when I was talking, and this was 2016, he had one is to 10,000 devices, one IT for 10,000 devices to manage. And he said, I'm looking in 2020 to be one IT for 250,000 devices, going up to a million devices. I'm like, dude, you're doing some funky math here. That looks like that hockey stick curve, right? And I'm like, he was, right now I don't even know what's on my network, what's connected to my network, I'm flying blind. And that opens up a lot of security issues, that opens up a lot of operational challenges. In fact, for every dollar our customer spends on CAPEX for buying the network, they spend $3 on OPEX, managing the network, monitoring and troubleshooting the network. So that's the key point, saying that you can hire 100 more IT staff, you're just not going to be able to manage the complexity. So there has to be an automation world, right? We live in a world where repetitive tasks should be done by machines and not human beings. It's happened in the rest of the lives and networks operations is just one part of that. So the concept of controller led architectures, which was the genesis of SDN is now being applied to this world of internet based networking, but we also get the data to provide you insights on how things are behaving and how to take actions before it happens. Well, yeah, you brought up, you used to, how many devices the enterprise can manage was something we measured for the longest time and used to compare to the hyperscalers. And I said, well, here's the myth there. It's not that they're managing two orders of magnitude more equipment, they architect completely different. Exactly. They build the applications with the expectation that everything underneath is going to change, it's going to fail, it's going to be upgraded. So you don't have somebody inside of Yahoo and Google and all these hyperscalers running around, patching and updating things. It's, they build a data center and they keep adding environments and they throw things out in the wood chipper when they're done and they break things down. So it's a completely different mindset and part of SDN was the promise of it was to take some of those hyperscaler methodologies and bring it to the enterprise. So yeah, tell us how your software today is delivering kind of that hyperscale architecture and it's a little bit of a culture change for the enterprise. It's been a huge culture change, right? Like the concepts of like abstracting the underlay complexity of all the network physical connections and giving you an overlay, what we call a fabric. So underlying network works as a single integrated system, right? It's not like switches, routers, controllers, access point, all of that complexity is taken out. So you're programming a single fabric, putting the right policy and the controller will figure out how do I enforce that policy in this switch, that place, this controller, this access point, right? So that was the complexity the network operators of yesteryears were dealing with, right? They had to go and configure with CLIs and now API, since we are in DevNet, is the new CLI, right? And that becomes a culture shift for network operators. Like I've been in the networking space for like 20 years. I was born on CLI, right? And even when I created systems like access control list, QOS, and I had to system test my own code, it's a freaking nightmare. It was tough, it was tough to manage that as a single system, right? And that's why the role of controller to abstract the complexity away, to program the infrastructure and then expose this intelligence to other systems, whether it's IT systems, whether it's business applications goes a long way. So that's why this journey is very exciting for us. So it sounds like we're entering the era of self-driving networks. I mean, you've got, I mean, to even visualize this is virtually impossible, unless it's at that abstraction layer. Yeah, yeah, absolutely. I mean, see, there are new technologies that a lot of consumer markets and other places have used, like machine learning, right? Like we have so much data within the network. The network sees everything, right? Because it's the connection point from mobile IoT to applications and cloud, right? But we haven't really leveraged the power of the data and the intelligence, right? And now that we have all of the data and now we have things like machine learning, it can identify traffic patterns and provide you more insights around your business, around your IT and security, right? So that really takes the guesswork away. And the good part is with machine learning, the more data you feed it, the more it's learning from the data, not just your own local networks, but the networks across the world. And that makes it constantly adapting to changing conditions and constantly learning based on the traffic patterns and your environment. And that's a pretty exciting field, right? Because we've implemented that in the security field to predict threats before they happen. We've implemented that in parts of application performance. And now you're bringing it to the world of networking across access branch, run and campus to like help IT move from a reactive world to more of a proactive world to a predictive world, right? So they can spend less time looking for the needle in the haystack and focus more on solving strategic problems. So when you get into discussions about machine intelligence, oftentimes there's discussions about, oh, replacing jobs and blah, blah, blah, blah. And so it'll turn to a discussion of augmented intelligence, which very reasonable thing. What you just described is removing mundane tasks, nobody wants to do those anymore. Here's my question. You talked about your CLI experience over the last 20 years. Is that CLI sort of tribal knowledge still vital as part of the art of networking or does the machine essentially take over and humans go on to other things? Yeah, I think that's a great question, Dave. I call these next generation of network operators the unicorns. So you do need to have the tribal knowledge of networking, not necessarily CLI, but the concept of networking, how do these protocols work, right? This is not easy. There are very, very few network engineers compared to application developers and software engineers in the world. So this is always going to be critical. But now if you marry this knowledge and compliment this knowledge with programmability and automation and application, you got yourself a unicorn that is going to be very, very strategic to the business because now the world of infrastructure and applications are coming together so you can truly focus on your business, which is run on applications, right? How can your applications run faster, smarter, better with the network? And how can your network understand how the applications are behaving becomes a whole new world? So you seek new roles of network practitioners emerging, I feel, like the data scientists of the network, like the security defender of the network, the world of security ops and networks are coming together. So that's what is exciting for us because you get bored in your life if you're doing just repetitive task and not learning new, right? And this provides a new way of learning. So for me, it's not taking jobs away. It's like upgrading your skillset to a whole new level that's a lot more exciting. This is the secret of Cisco still. We've talked about this all these hundreds of thousands of network engineers with a growth path and come develop. What I found fascinating is really unlocking that data because for the last decade, we've talked about, well, there's the network flows and there's analytics in the network streams but what had been missing and what I think is starting to be there, as you said, that connectivity between the application and the actual data for the business, it isn't just some arcane dark art of networking and we're making that run faster, better, cheaper but what that enables for the business, the data and the applications that there is a tighter relevance there today. That's the key thing, right? I mean, everybody has been talking about data now, I don't know, for 15, 20 years. It's the new crude oil, if you will, right? But everybody has access to data, nobody knows what to do with it, right? Like this philosophical thing of data to knowledge, to wisdom is like what we are all striving towards, right? And now that we have access to this data and we have this intelligent system, which is a multi-purpose software that ingests data from not just networking but devices connected to the network, the security threats that we are seeing, the application data that we're seeing and provides this context and provides you very key insights around how does that impact your business? How does that impact your IT? How does that impact your security? It's a very powerful thing and you don't find that and you need to have that breadth of portfolio and system to be able to get all of the data and consume that at a hyperscale level, if you will. Yeah, and we often say in theCUBE that data is plentiful, insights are not and you need insights in order to be able to take action and that's where automation comes in. Prasad, great segment, thank you very much for coming on theCUBE, really appreciate it. Thank you, Dave. Thanks, Stu. Pleasure. Awesome. All right, and thank you for watching. This is theCUBE live from Barcelona Cisco Live 2020. Dave Vellante for Stu Miniman and John Furrier. We'll be right back.