 Live from Barcelona, Spain, it's theCUBE! Covering Cisco Live Europe. Brought to you by Cisco and its ecosystem partners. Hello everyone, welcome back to theCUBE's live coverage here in Barcelona, Spain for Cisco Live Europe 2019. I'm John Furrier, your host of theCUBE with Dave Vellante, as well as Stu Miniman, who's been doing interviews with us all week. Our next guest is Roland Acura, Senior Vice President, General Manager of the Data Center Group. He's in charge of that core business, the data center now at the center of cloud and Edge, Roland, great to see you. Thanks for coming on. Thank you, thank you for having me. So, a lot of announcements. So all the big guns are out there for Cisco. You got the data center, you got the networking group, and you got IoT, and then cloud center suite was part of the big announcement. Your team had a big piece of the keynote yesterday and continues to make waves. Give us a quick update on the news, the key points. What was the announcements? Yeah, the two big announcements from my group were ACI Anywhere and Hyperflex Anywhere, and we captured them under a common moniker of there's nothing centered about the data center anymore, because both of these speak to things going outside the data center. ACI Anywhere is the integration of ACI, our software defined networking solution, into the two of the most prominent public cloud providers out there, Amazon and Azure. And for Hyperflex Anywhere, the exciting news is the expansion of Hyperflex, which is our hyper-converged solution, also outside the data center to the edge of the enterprise, specifically branch offices and remote locations. And the other thing that came out of our conversations here on theCUBE and also on the keynote is that the center of the value is the data center, as you guys pointed out with the slides, big circle in the middle, ACI Anywhere, Hyperflex Anywhere. But the network and the data and the security foundation has been a critical part of this new growth. Take a minute to explain the journey of ACI, how it started, where are we? It's been a progression for you guys, certainly inside the enterprise. But now it's extended. What's the journey, take us through that. Yeah, when ACI came into the market five years ago now, we have a five year anniversary, ACI brought a software defined networking solution into the market. It brought an automated network fabric capability which said you can no longer screw yourself up by having incoherence between one part of the network or another, it's all managed coherently as one thing. And it brought to your point about security what's called segmentation of applications. Today applications have data, they have databases, they have different sensitive pieces and it's important to be able to tell the network not only get the traffic from one place to the other but also selectively get the traffic that I tell you to get there and not the one and don't get there the traffic that has no business getting there. And that's known as segmentation which is a security concern. Particularly when you have sensitive data like consumer data or things that have regulatory things around them, ACI has brought that to the market. That was the value proposition of ACI. We worked on then expanding ACI in the direction of scale. Customers have two or more data centers for disaster recovery for resiliency. We made that possible. We got to bigger and bigger footprints. Then we took ACI to the edge of the enterprise. What if somebody wanted to put some computing capability in a store or in a logistics center? ACI then was expanded with that. Step N minus one was we took ACI to bare metal clouds. Customers now want to deploy also things in co-locations or bare metal clouds. We decoupled ACI software from the Cisco switches which is the ACI hardware and ACI became completely virtualized and still able to be doing everything it does in hardware on premise in software instead in somebody else's capability. And yesterday we announced the full culmination of this which is what if there is, you don't want the ACI soft switching or hard switching. Can you use the native switching of a public cloud like Azure or AWS and you tell them the other APIs, please let those packets go from A to B because they're part of the white listed paths and don't let packets from C to D go because they're part of the black listed paths. And that was the full integration with these clouds. And you abstract that complexity? Completely, completely one orchestrator which is the multi-site orchestrator. The same one people have used on premise that they've developed their policies around so that we have invested a lot of sweat equity in that controller. It's where also they put their compliance, verification and audit and assurance and they use that thing even when something goes to Azure or it goes to AWS. So you mentioned the progression. So this is now your full progression from core to the cloud, including Edge. Going through Edge. What has been some of the results? You mentioned that segmentations with one of them. I get that. How has ACI been used? What are some highlights that show the value because people are looking at ACI saying, hmm, I like this. I like scale. I have a scale challenge with the new cloud world and Edge and complexity is abstracted away with software. Okay, check so far so good. Where has been the success of ACI and how do you see that unfolding specifically in the cloud? Yeah, the biggest value our customers have gotten cloud or no cloud has been with ACI. They've been able to shorten the speed of change, shorten the time for change, therefore increase the speed of change of their network because now the network needs to operate at the speed of the applications. Applications reconfigure themselves sometimes on hourly or daily basis. And it used to be that changing something in the network, you know, you sent a ticket to somebody who took weeks to reconfigure things. Now that software defined capability means the network reconfigures and people can change generations of compute on the fly and the network is in lockstep with them. The agility and speed has been great. The other value has been the value of automation, which is people can run a bigger and bigger and bigger network with a small number of people. You don't have to scale your people the more switches you have. Again, because programming and automation comes to the rescue. Well, I'll tell you, people who I'm watching right now can look behind Roland and see that it's packed house. We're in the DevNet zone, which has been the massively growing organization within Cisco community has been growing very fast. People are developing on top of the networks and these are network folks and it was new talent coming in. So skill gap is shortening. So he's getting a different makeup for a Cisco user. Your customers are changing and changing growing existing base plus new people. How about that dynamic and how that impacts this intent-based networking, this notion of policy and software to find? What many people have been calling infrastructure as code, which is you go from scripting to actually coding and composing very sophisticated automation capabilities and change management capabilities for an automatable system, which is what ACI is. It's made for people drawing on the strengths that they were doing in the application domain or in the server domain and bringing that into the network. And that's a new and exciting things. It brought the network within the purview of coders of people who know how to do Python or know how to do Go language and things which are modern and exciting for the younger generation. It's made also for bringing the analytical capabilities. A lot of what those young coders are used to is a lot of logs, a lot of visibility, a lot of analytics running on, because they've done that on web servers, they've done that on applications that run in the cloud and we now offer the network, which is very rich in data. If you think about, we see every packet, we see every flow, we see every pattern of how the traffic is changing and that becomes a data set that is subject to programming because then from there you can extract anomaly detection, you can extract security signatures of malware, you could extract prediction of where the traffic is going to be going in six months. There's a lot of exciting potential from the telemetry and the visibility that we bring into that framework. And as you pointed out, Devs love that. I mean, Cisco, we've talked about this as one of the few large established companies that has, in our view, figured out developers. There's a lot of examples of those companies that have it and continue to struggle, but just witnessed here, the Dev Crowd. I want to ask you about ACI and how it's different from, for example, VMware NSX. What's the differentiation there? Yeah, the biggest differentiation is ACI is one system through which you manage the entire network, the overlay, which is the virtual view of the network that the applications care about, as well as the underlay, which is the actual real delivery system that makes the packets get from A to B with quality of service and so forth. So that's the first thing. It actually does a lot more. It has much more scope than NSX does. The other thing that's very unique about ACI is we have integrated it with every hypervisor on the planet and every container management framework on the planet and every bare metal system on the planet, which means that any workload, something sitting on a mainframe, something sitting on a Sun Oracle server, something on OpenStack, on OpenShift, on VMware, on Hyper-V, and now on the EC2 APIs of AWS or on Azure, all of those are integrated with ACI. We're not wedded to one hypervisor and our cloud implementation that we announced yesterday is a true integrated cloud capability. It's not a bring your own license and go put it on bare metal at AWS, which has been VMware's cloud strategy is to team up with AWS and let customers bring their software licenses into AWS bare metal. That's not EC2. And of course that's not Azure and that's not the other clouds we're going to be doing. So the openness to being multicloud on premise, which means every hypervisor and every container framework and bare metal with one system, we're extending that into the cloud to give customers choice and openness. That's really a very fundamental philosophy. So much wider scope. And it's not always been Cisco's philosophy or partnership. When you think about HyperFlex, go back 10 years when you guys sort of created that with partners and then multiple partners now, maybe talk about that journey a little bit. HyperFlex? Yes. Yeah, because hyperconversions is another very exciting and fast growing trend in our industry. And really HyperFlex started off and with the hyperconverse infrastructure, started off being the notion of putting a mini cloud in a box on premise for application developers to rapidly deploy their applications as if it was in the cloud. So speed and simplicity were really at a premium and that's really what defines hyperconvergence. And we've done a tremendous amount of work at Cisco to make speed and simplicity there because we've integrated network compute storage and a cloud management system called InterSight to give that whole capability to customers. We then hardened it. We took it from being able to do VDI kind of workloads and rather benign workloads to mission critical workloads. So databases are now running on HyperFlex. ERP systems are running on HyperFlex. The real crown jewels of the enterprise are now running on HyperFlex. Then we made it multi-cloud. We opened it to all hypervisors and to all container frameworks. We announced OpenShift yesterday. We had already done HyperV. We had done OpenStack and ESX. So again, same spirit of openness. And yesterday's announcement was, what if I want to take hyperconvergence outside of the data center in hundreds or thousands of remote locations? Think a retailer, okay? In a retail environment, some of the most interesting data is born outside the data center. It's born in a store. The data is the sensor that follows the customer who's interested in a plasma TV. And that data has a perishable lifetime. You act on it on location and on time, or you lose the value, right? So sending it over, taking two hours to do a machine learning job on it and come back, the customer's already back home watching a movie. And so the window of opportunity for the data is often right there and then. And that's why our customers are taking their computing environment often to where the data is to act on it fast and in the on location. It sounds easy, but I want to just get your thoughts on this, because this is a critical data challenge. If data is stored in classic old ways, data warehouses and fenced off area, it's kind of in the internet, you're not going to have the latency to get that data in real time. Talk about real time data that's addressable. Or part of the application value. So this is a new notion that's emerged with DevOps and infrastructure as code. That's right. That's hard. How do you guys see that progressing? How should customers prepare to have that data centered properly for app addressability, discovery, whatever the usage of the data contextually is, time series data or whatever data it is. This is a critical thing. It's a critical thing and there's no one answer because depending on what the data is, sometimes you only see the value when you concentrate it and consolidate it because the patterns emerge from rolling up a thousand stores worth of data and seeing that people who buy this toothbrush and to buy that toothpaste, that maybe that value which is where you want to concentrate the data. But there are also many things where acting on the data in the moment and on location quickly without referring to the other thousand stores extracts 90% of the value of that data. So that's why you want to do forward deployed computing on that data. So this highlights network programmability. This means the applications driving the queries or the network for that data if it's available. So there's two things. Network programmability from the app and availability of the data. Yeah, and the ability for the entire infrastructure, network compute and storage and hyperconvergence is the automation of all three to be able to deliver its value equally in remote locations or in a cloud as it would have in a data center because that's where the application is going to want to go where the value is and if the infrastructure can't follow it there then you get a degraded ability to take advantage of the opportunity. Right, real-time decisions happen at the edge but then as you described you got to bring data back, certain data back to the cloud, do the modeling there and then push the models back down. So you're going to have- Decision making distributed. And you got to have low latency to be able to enable that. And the same goes for other considerations. For example, why is it important to allow people to put data both on their premises and in the cloud for disaster recovery, for data replication, for resiliency, sometimes for governance reasons. GDPR in Europe says the data of European citizens that's personally identifying has to stay in Europe. Somebody may not have a data center in Europe. Can I take advantage of a co-location ability or somebody else's? This is the theme we're seeing at this show this year and certainly at the center of the news is complexity is increasing because there's just evolution. More devices are connected, diverse environments, scale for cloud and connectivity but software driving that. So I got to ask you the question, go back to the old days in the 1990s, multi-vendor was a big word. Now multi-cloud feels the same way. Yes. This is the openness thing. How would you describe multi-cloud strategy for Cisco in context of this notion of being open? It is really the new dimension of openness, right? We've been open in the past to multiple forms of physical networks. Customers to use wireless or fiber or copper or what have you. We need to give them an IP network that operated equally well over all media. That was one dimension of openness. Another dimension of openness was does a product from vendor A work with product from vendor B? My router, your router, my switch, your firewall, those are other dimensions. Hardware and software coupling. Can I buy the hardware from Peter and the software from Mary? Will it work well? The new dimension of openness is can a customer avail themselves of any form of cloud either because they like the tooling and how well their developers are more efficient on a given cloud or because the pricing of the other guy or the third guy has a point of presence in Tokyo which this one doesn't. All of those are business choices that if we make our technology, let them take advantage of them with no technical restriction, they win. Because now they can shop on the merits of what they want to do. And not on, oh, I'm sorry, if you want to go to Azure, I can't help you. But if you're willing to settle for your own premise or for Amazon, then I have a story for you. So that's- Rowan, you're leading the team on the core crown jewels for Cisco. As you guys, the rising tide is floating all boats here within the company. What's your plan for the year? What's your goals? You'd be out there pounding the pavement with customers. What's your objective? What do you hope to accomplish this year in 2019? Well, 2019 is the year of many things for us. It's a very exciting year. It's the year of, on the physical infrastructure side, we're taking our switches to 400 gigabit per second. We have our new silicon capability, our new optics. So we're going to be able to scale for the cloud providers who are hitting the next frontier of speed and density and scale. So performance will always, always be there. And when we're done with 400, we're already going to be asked about 800. So that's an exciting new generation of switches. ACI, anywhere getting deployed now and adopted across multiple clouds is another exciting thing. Hyperflex, anywhere, we're really looking forward to the potential in financial services, in logistics, in retail, where there's a lot of deployed data at the edge. And then security is a never-finished journey, right? Everything we give our customers in the way of security, because there's an active actor who's trying to make you fail, right? It's not that you're only fighting physics to get to 400 gigabit, then you win. There you have a guy who's trying to foil your schemes and trying to foil their schemes. Constant attacks are on the network. You guys have seen this movie before, so you know how critical it is. Roland, thanks so much for spending the time. Congratulations on ACI, anywhere, Hyperflex, anywhere. Intent-based networking at the core. It's theCUBE bringing you all the data. We have an intent here to bring you the best content from Cisco Live in Barcelona. I'm John Furrier, Dave Vellante. Stay with us for more live coverage. Day two of three days of coverage, here in the DevNet Zone, packed with developers. Learn on new skills. We'll be back with more after this short break. Thanks.