 Live from Barcelona, Spain. It's theCUBE, covering Cisco Live 2020. Brought to you by Cisco and its ecosystem partners. Welcome back. This is theCUBE's live coverage of Cisco Live 2020. Here in Barcelona, it's our second full day of coverage. We're actually doing about three and a little bit more days of coverage. I'm Stu Miniman, I co-host for this segment is Dave Vellante. John Furrier is also in the house. A lot of interesting announcements at Cisco. We've been watching for the three years we've been at Cisco Live, really Cisco getting much deeper into the software space. Of course, we're here in the DevNet Zone where we're watching the changing workforce go even more towards developers, hardware and software are living together nicely. And to help us dig into that topic, welcome back one of our CUBE alumni, Todd Brannon, who is the senior director of computing with Cisco. Todd, thanks so much for joining us. It's a pleasure being here with you. Good to see you, man. So let's tee up what I was just talking about there. You know, there are certain companies that think about Cisco in the boxes and ports. We know the future is more, you know, software is eating the world, developers of course are the new king makers. That's right. And Cisco's been moving on that journey, so bring us inside a little bit the announcements that you've been working on and you know, where are your teams seeing where customers are going? It's all about the application. So the app dynamics team, they did some surveys. They found that the average consumers, 7,000 people worldwide, they surveyed them. Average consumers using 34 apps a day, or 34 digital, you know, experiences. And so if you think about applications and infrastructure, we've always talked about applications and infrastructure, right? People don't buy a server to use as a coat warmer. It's always going to be running some sort of workload, but I think in the past, there wasn't as visceral connection between what the operations and infrastructure teams were running. If it was CRM or ERP, it wasn't as visceral connection as you have today when it's say the hotels interface to their customers in terms of booking or getting that early checkout, right? So the application experience has become much more personal, much more visceral and really incredibly critical to the business. And so that puts enormous pressure on the infrastructure teams that are a big part of making sure that application experience is a good one, right? Response timer is the app even available. So for us, it's how do we start to begin to bring that infrastructure operations team into the fight with the teams that are thinking about applications, which have oftentimes been using different tool sets or sort of some operational silos though. So this announcement's all about trying to break down some of those silos all in service of that application experience. Right, and of course, AppDynamics was a large bet by Cisco. We've seen them in a lot of the cloud environment, very much tied to the application. One of the announcements this week is taking Cisco InterSight, which the first time I ran across it was in things like UCS and other Cisco gear there. Help us understand how those are going together now and how should we be thinking about InterSight today in 2020? We should definitely be thinking about it differently because you're right, here too for InterSight has been focused on our computing infrastructure, HyperFlex, UCS. And when we started with UCS, we took management out of devices, we moved it into the network fabric. And then three years ago with InterSight, we moved our management control plane into the cloud. So think of Meraki, but for computing. But it was always around Cisco and our infrastructure. Now we're taking two really big steps. One is we're integrating a product that we've had called our workload optimizer into InterSight and that workload optimizer software has always been inherently heterogeneous approach. So databases, cloud management platforms, all the hypervisors operating systems, storage partnerships, we've been able to do telemetry and interdependency mapping with that and a heterogeneous way for some time. So now it breaks InterSight out of being more of a Cisco focus to really the reality which is a heterogeneous data center. Second thing is that now we've done a data integration with AppD. And so the beauty of that is AppD best in class at understanding the interdependencies, those complex web of interdependencies that the application tier, our workload optimizer best in class, understanding infrastructure interdependencies and now we correlate those and you get a top to bottom view. Again, that kind of gets you to where you can see what's an application, how's it performing from a business context with AppD but been able to connect it all the way down into your cloud and on-prem infrastructure, make that correlation and ensure that your infrastructure is doing the right things for the app. So the application portfolio has evolved dramatically over the last 20 years, right? It used to be, and it still is, the crown jewels of the organization but you'd have a mission critical app that the insurance company would have a claims app and then there'd be a zillion other applications around it but the claims and the sales apps were really the key and whatever happened in the other apps, okay, fine. Now you have developers, you have shadow IT came in and Stu's saying everybody's a software company and you had this explosion of apps that are all sucking resources from the network so the traffic has changed and that brings massive complexity. So I wonder if you could talk about how that trend has affected network traffic. You hit it, it's the interdependency. So it's been estimated that the typical enterprise application until very recently needed to talk to four to eight other enterprise applications to function properly. We're seeing that number jump to 20 in a very near future and it's because applications to your point are becoming much more modular, right? The development environment in the cloud where all the innovation is occurring is inherently distributed, microservices, serverless or functions based in many cases. So all of the things that conspire together to create that experience like an insurance claim on your phone, all of those interdependent components it's become much more distributed, much more complex and the key thing is underneath each of those components is going to sometimes be different infrastructure managed by different teams with different tooling and so it's become almost impossible for IT teams to correlate all that and manage it, especially as it becomes higher velocity. Right, God's do it. Yeah, Todd, would like you to put a point on that because you've talked about applications are becoming much more distributed. Want to hear what you're hearing from customers because sometimes it's like, well, I think of this application as either one thing or this collection of things that I put one place or another. We're starting to see some customers that, well, I start teasing things apart and therefore it becomes developed hybrid in nature. Exactly. And people often complain it's hybrid, it's multi, it's all this other things. Well, it's a real difference between I had something in my data center and a piece of it is in a public cloud versus, oh, hey, I'm just going to throw thing in whatever public cloud I want to use today or tomorrow. I think that's an incredibly important distinction. So multi cloud is the notion of, hey, I want to be able to consume innovation from different cloud providers, but a hybrid application is really this idea of public cloud. So microservice connecting back to monolith on the on-prem side. And I think it's still very, very rare that people are building applications that tie together multiple public cloud services to your point, but it's very much more common for people to be saying, hey, I've gone out and built something innovative, a new customer experience out in the public cloud, but now I have to connect it, data gravity's real. GDPR here in Europe, right? So there are very real reasons why applications and data are staying on-prem, but they need to connect it out to this cloud innovation, and that's what this announcement was all about. How do we give people a tool set? Because if you think about it, you're going to have infrastructure powering these pieces in the cloud on-prem. How do you monitor that? How do you ensure that you're not over-provisioning or under-provisioning? It's a very complex problem. Well, it's critical because the cloud brings scale. It used to be, okay, we're going to deploy a website. Hey, the website's important. It's slow. Let's figure it out. Now you have these dozens and hundreds of applications coming in, many if not most of which are customer-facing, so if there's a problem, it's really escalated and the cloud helps scale that problem massively. So, Todd, help us understand sort of, in the keynote yesterday, there was a sort of a circular diagram, the visualization, the insight, and the action. So give us a little sort of insight as to how this works. What's the secret sauce underneath it? So the secret sauce is correlating data, right? So telemetry data is something that we've always collected in the context of either infrastructure or applications. So with AppDynamics, we have a platform that best in the industry at going out and figuring out all the interdependencies between an application and all of those services that are there. And then we have all of the similar things on the infrastructure side. And so what we've done here is correlate those data sets. So we're using the API as a feed data between AppDynamics and Cisco Intersite, which is the infrastructure side of the equation, and we create a data lake now that we can then be able to apply analytics to. And so we can start to think about the data center as a demand supply equation. And how do I want to match up my applications with the business context and tact from AppD to what I'm doing with my infrastructure and provisioning that. So it's really a story of collecting all the telemetry, integrating it, stitching it together, and then applying the analytics to help our operators because it's gone beyond human scale, keeping track of the needs of all these VMs, and especially when you get to containers. So it's first about stitching together the data, then applying the analytics for insight, and then taking action. So it's automation informed by insight, but first you have to have visibility of everything. So that's the loop. It's interesting you're talking about demand supply. Again, it used to be, you'd manage demand, IT demand, with an IT project management system. And now you've got this infrastructure that is being sucking, apps are sucking resources out of it, and you can't just manage it manually. You've got to have the data, which you've got, and you've got to have some level of automation to be able to remediate things. So how does that fit in to the product and sort of the roadmap? So our Optimizer product has, you're going to give your credentials for all of the different tooling in your data center, and you're going to bring it all together for the analytics and then be able to take action in a similar fashion from a central position. So what you see in InterSight Optimizer is really powerful as a recommendation engine. So it's going to tell you straight up, hey, you've got an, you have an application, and it's going to look at historical data. So over the past whatever, 30 days, this VM over on AWS, 95% of the time has been running at less than 70% utilization of its assigned resources. So guess what? You should go from instance size three to instance size two, and we can even tell the operator, here's how much money that's going to save you every month. Do you want to do this? Yes or no? Bang, off you go and kind of stand up that new instance. Similarly on the on-prem side, this VM has been consuming more than 95% of its allocated memory, 80% of the time over the past month, you should give it some more memory. And because we have Optimizer controlling vCenter or the mic, we can go off and make that change. So it's really the analytics to decide what is the right action to take, then given the operator the go button to go instantiate it, and that's incredibly powerful. And it's the same experience for my on-prem workloads, my Amazon, my Azure, my Alibaba, whatever, the workloads I'm going to run in the future. Correct, and that's essential because of the hybrid dynamic. The innovation's going to go on out in the cloud, but you've got to tie at the back ends. We have to be able to manage both of these at the same time. So people might be asking that aren't as into this world as well, why can't I just, isn't Amazon going to do that for me? Isn't Azure going to do that for me? Or the IBM cloud, whatever. Can you explain sort of help people understand the differences in the way in which each of these environments, including on-prem, handles this type of activity? I think what we're seeing is a maturation of the on-prem side of the equation. So the cloud-like operating model, consuming resources, that model, that cloud's an operating model, it's not a place, right? Everyone's been throwing that and we're out for a few years, but it's very true. And so now on-prem, OpenStack was hard for folks. We know that it just was difficult for people to get to the private cloud, the nirvana that they wanted to. So with things like InterSight, we're basically starting to deliver enterprise-ready, hardened systems. We're not calling it a private cloud, but effectively that's what it is. Especially when we talk tomorrow about HXAP and what we're doing on the container side, that's ultimately what we're delivering is a cloud-like experience for the operator. So as a company, we've been focused for 10 years on how do we create a better operating model in the data center, but now we're competing on experience just like our customers are with their app. So we have a mobile app for InterSight, right? And we're focused now on the experience for the operator and bringing that cloud-like experience on-prem. That's really the... Todd, I'd like you to dig into the organizational impact here a little bit. First of all, from your partner selling these solutions into the customer, as well as from a customer standpoint, because I kind of hear and visualize a little bit, well, you know, AppD is very much an application-centric focus, as opposed to InterSight is more the infrastructure piece of it and those worlds haven't necessarily communicated or there's some gaps. They have been the victim of silos on a technology basis and then that does manifest in the organization, right? And we used to see this when we started with blades back in the early aughts, right? Is the networking team going to sign off on this blade chassis? Well, they can't manage the switching. We're not going to let the server guys manage that, right? Technology kind of reveal very dysfunctional organizational constructs. And I think we're trying to help the same dynamic here, but between the folks that are concerned about the application and how it relates to the business and looking at the application performance and then teams that manage infrastructure, they haven't had common tooling. And this provides common data sets, a single source of truth, so that when something goes wrong, everyone's aware of the same set of conditions they can see, they can correlate it. We're correlating these two data sets from the app side and the infrastructure side and it helps the teams work together. Because you're right, I mean, you've got app teams that look at the world as a, you think of it as a horizontal application topology, but underneath every one of those points on the graph, there's an infrastructure component, maybe different teams. So, and they're looking at the world as stacks. You got the infrastructure folks looking up, the app folks looking down, unless you've got these worlds correlated, that's what the war rooms and the finger pointing, it must be the network, no, it's, you know, so we're really trying to help teams come together because ultimately in a business, they're all working for somebody that cares about the whole enchilada. Okay, so from a selling motion, is it that person that they're report up to that will drive that? It's both. Well, what we're doing is, you know, we have our infrastructure operations teams, the folks that we work with there. Now we can bring them a tool set that says, here's how we can help you be directly relevant to the business in real time. Here's how to hug your application team and make them happy, right? So it's a story of relevance and in a real time way. And then for the application team, it's a story of, hey, here's a tool set that ensures the thing that you care about most, which is your precious baby, your application, is going to get all the care and feeding it needs from the infrastructure on-prem and the cloud. And so our app D team is talking to those application-centric monitoring and operations teams and all the folks that work in our data center or organization are talking to the infrastructure buyers, but we're now selling them a common tool set. You know, one team kind of come and bottom up, the other come and top down. And it's heterogeneous, I don't need, I don't have to have Cisco gear to make this work. And it's a SaaS model. It is SaaS, it's crucial. And it's a 2020 availability, right? The calendar 2020. Yes, first half. First half, great. All right, Todd, want to give you the final word as we look through 2020, what should be customers be looking for in this space? I think they should be thinking about how can they impact the top line and the bottom line? So as an IT organization, on the top line, it's going to be these new application experiences. That's where the companies are innovating, right? To drive revenue new experiences. And then on the bottom line is, how do we get rid of over-provisioning? How do we operate in a more efficient way? And to do that, you need analytics, right? I haven't said AI ops, but I'll throw it out in the clothes, right? But you need analytics to really understand how do I optimize the environment, reduce my cost of computing, and help out with the bottom line. So that's the rest of the year. Todd Brennan, really appreciate the conversation. Thanks so much for all the updates. Look forward to talking to you again soon. Thank you, pleasure to be here. All right, for Dave Vellante, I'm Stu Miniman, back with much more wall-to-wall coverage here from Cisco Live 2020 in Barcelona. Thank you for watching theCUBE.