 Okay, we're back here live at HP Discover. This is, we're in Las Vegas for day three of three days of wall-to-wall coverage at HP Discover. This is theCUBE, our flagship program. We go out to the advanced instructor, Sillen from The Noise. I'm John Furrier, the founder of SiliconANGLE, and I'm joined by Koos. Hi everybody, I'm Dave Vellante at wikibon.org. And this is theCUBE. We're here with Kristoff Fister, who's the vice president and general manager of the business service management component of HP software. Kristoff, welcome to theCUBE. Thank you so much. So tell us a little bit about the business that you run within HP software. What's the focus? Absolutely, so it's called business service management, and what we do is we ensure availability and performance of large data centers, the infrastructure, as well as the applications that run within that infrastructure. So we're talking about problem management, change management, is that right, or? It's more identification, event management, the analytics that go with it, and across all components of the data center, networks, systems, storage, the apps, and everything around it. And does helpdesk fit into that, or no? Helpdesk is closely connected to it, so people typically deploy our solutions in what is called the closed loop incident problem management framework. And so once they get events and they need to deal with them in a more systemic manner, meaning there's no quick fix, it becomes a trouble ticket, and that's when it goes into the helpdesk and the change management system. And once that's done, it comes back to us and the events get closed out, so it's very closely connected. Ah, okay, so the change management database is part of your organization, or is it not? It's not, okay, great. So you have to feed that, yes. But you work across these boundaries. Okay, so you guys have the application lifecycle type part of the software group, is that your group too? So look at it as the application lifecycle management to the left of me, that's where the development happens. Once the development of the app is done, it gets into production, that's where the operations people are, and they use business service management tools. And then once things break bad enough, then it goes through the service management. So I see it in the middle between these two. So you get to fight the ops and the dev guys. So you're the dev ops broker, right? So you're brokering no downtime to, hey, I'm just iterating fast, right? So that's the challenge we're living in right now is the dev ops culture is hitting the enterprise. That kind of is... Yeah, so, you know, there's sort of the dev ops religion, and then there's what enterprises experience. And it's true that there's still significant balls between the development folks and the operations. And yeah, some of the stuff we're trying to do is trying to break down these balls. Yeah, IT service management is still quite siloed. And so talk about how you're breaking those down. I mean, one way is obviously you can have a single CMDB, but a lot of times customers don't want to go to a single CMDB because they have their own database. So talk about what's going on. What are the big trends that you're seeing within service management? Yeah, absolutely. So one of the things we're hearing a lot now is the old management paradigm of sort of instrumenting your IT environment for known problems. Where you have the systems, you say, I got to drop an agent on that thing and then they're going to instrument that agent for events and all that stuff. It's becoming difficult to do. Why? Because these IT environments are becoming so dynamic. They're growing so fast. And so this paradigm of management by exception is something that customers are more and more struggling with. And so we see a shift happening where instead of that paradigm, there's this notion of collect, store, and analyze that's becoming a trend. And so that's exactly what we're responding to with this notion of operational analytics, where you bring all this machine data into one place and then you drive analytics on top. So one of the things we had George Kedifon yesterday, EVP of the software group, and we were talking about the new style of business, not just IT, but business, and that's certification. I mean, I'm oversimplifying kind of what, now it's my words, but he was basically saying certification, new ways to write code, and he made a reference that with open source, there are now a variety of different languages to program in. So one of the hot areas that we didn't get a chance to talk about is visualization. So on big data, that's a very, very hot topic, visualizing data. So on the code side, you guys have a lot of SaaS offering where developers can actually look at visualization, a lot of automation on code. Can you just give us an update on where that is in this operational analytics piece? Yeah, absolutely. So we were focused on doing advanced analytics on top of this data lake, if you want to call it that way. And advanced analytics has a number of different components. It has a component of machine learning, algorithms, but also visualization is becoming very, very key. Why? Because to be able to understand what's going on, humans can, I read that once we found that interesting, in text form humans can process about 200 words per minute. If you give them a picture, they can process the equivalent of about 2000 words per minute. So visual is, you know, makes business sense. It's not just, you know... Make sure it's worth a thousand words, literally thousands of words. Exactly. We're 2,000 words. Yeah, but that's true. We've got the real stats to back it up now. And so visual analytics is part of what we do around advanced analytics, and so... Good, sorry. Yeah, I... That was your demo yesterday, right? Yeah. That's why I wanted to connect to that. Or you use, you know, hot data type of charts. And that, you know, that allows you to represent heat in the IT environment. It's very, very critical. And a big complex in the IT environment, you've earned my hotspots. And it's using, it's not just the same technologies that are applied in... You know, I'm from Germany, and Germans are big on, you know, green. Yeah. And so they do... Americans, too. You know, pointing thermal cameras at their houses to see where the heat escapes. You know, it's using some of these same, same ideas in the same context. Point that camera at my house. But so, but for those of you who didn't see the demo, it was, it was very cool. It's like a heat map, but it's also a time machine. So you gave the demo, right, on stage yesterday. So you basically were able to dial back. So you said, you saw red, bad. And then you were able to dial back in very granular increments. I mean, by the minute or second or whatever it was, until you started to see, it's almost like a weather map when you see the, you know, the storms coming in. Is it? It's red, red, red. No, it's good. Oh, what happened now? Okay, so how do you determine what happened? How does an individual determine what happened? Well, so, you know, that this time lapse analytics is one of the analytics we do. One of the advanced analytics. And it's really the concept of, you have all this data in one place. You collect it over, you know, long periods of time. And you have it in these high speed databases. And so all of a sudden you can go, you know, back and forward in time in your IT environment, because the speed is there. You have the raw, you know, processing power to do that kind of stuff. Before, you know, in these BI type solutions, you could do similar things, but you had to type a query, wait for six hours, look at it, and then, you know, sort of think about what you need to do next. Now you can go, you know, back in, you know, whatever increment you want. So you can go back and look at your IT environment, how it was three hours ago. And if it wasn't better shaped than it is now. And then what might have, you know, potentially caused the issue because you don't have just the performance metric and the visuals, you also have the log data at exactly that point in time. And so it's very, very quick to sort of pinpoint issues. And it's very exciting. So this time lapse analytics, this is new, obviously. Absolutely new. So how do I buy it? If I have a license of your existing IT service management, can I add onto this? Not as a consumer. So we're announcing it at this show and it's going to be available in Q3 calendar of this year. So you can't buy it right now. Right, okay. But once it's available, you can absolutely buy a license and deploy it and get better. What's happening in terms of the sassification of your business? Because a lot of IT service management products are, you know, installed on-premise. You know, many of them are many decades old and they've evolved over time. But is that model moving to SaaS and what's HP doing in that regard? Yeah, so we see, obviously the help desk market is the one that's furthest along in terms of sassification. You know, we're probably at 70% or so penetration already. Other areas of ITSM or business service management are a little bit further behind because, you know, especially on the infrastructure side, it's not, you know, you have agents to deal with. And so it's not, it doesn't suit itself as well for SaaS. But we think the next piece that's going to go is around application performance management. Where you do end user monitoring, end user behavior type stuff. And so we brought out about a month ago what we call our performance anywhere solution which is application performance management on SaaS. And it brings together, you know, synthetic monitoring which is sort of, you know, the robots go out and test performance against the website with, you know, more deeper, really user monitoring where you really get, you know, the experience that a single user has and we do that by, you know, sniffing out the transactions and figuring out what the response time is for a particular reason. So this is, you know, the next big thing in SaaS, we think. And one last question, if I may, before I turn over to John. So are customers still using a lot of spreadsheets to do the sort of infrastructure, incident management, problem management, or is it really beyond that? No, we're beyond that. But, you know, where people are, I was in Japan two weeks ago and what really struck me there is that, you know, people are still very, very focused on, you know, the hardcore infrastructure management, managing networks, managing systems. And we think, you know, the time is now to look beyond that and look at, you know, what's the end user experience on top of what the infrastructure monitoring provides? Because very often times your infrastructure looks green, yet, you know, the users are calling up and saying, hey, I got the performance that sucks. And so you want to get ahead of that game. And so that's why the combination of application performance top down and bottom up infrastructure management is such a big deal. I mean, I saw obviously the great story from HP. Obviously they have a lot of perspective with the big data from the network and the app side. But I want to ask you about the market trends. So obviously for applications, automation is very important, right? So, but with open source, social presence, there are people who are located all around the world. So talk about the collaboration development environment that's there with open source and talk about some of the automation and how do you guys address that? And specifically around pricing and business model, because it's clear that, you know, your vision is right on the money with, you know, the idea of how people develop, right? It develops all around the world, follow the sun, variety of different perspectives. Yeah, we do that ourselves. And also automation. George was mentioning yesterday, Kedifu was like, okay, everything was green, but also there's one spot of the code that might not be seen. We want a predictive mechanism to fix those things. So that's automation and that's, so developers need that or they need someone else who might be coding to see the changes. So there's all this change management of the code to automation. So talk about the productivity, the social collaboration and then how do you guys price that? Remember, I'm on the operations side of the house. So, you know, the development piece is not really, really mine, but, you know, we see a lot of, you know, automation happening there in the sense that, you know, once you discover a problem and you can use analytics to do that and do it very, very quickly with some of the new technologies we're bringing out. But then you would also, you know, very quickly remediate the issue. And their automation is a big deal now in terms of, you know, instead of an operator going, you know, through their own scripts and all that kind of stuff, you can automate a lot of, a lot of through aviation. The way we do it is through runbooks. So we integrate runbooks into our monitoring solution. And so these runbooks then can either run fully automatically if that's what the customer desires or the operator can step through, you know, in a step-by-step way through the runbook that at any given point in time say, you know, I want to execute the step or I don't. And so that's how we look at automation in the context of monitoring. Let's talk operations then because since that's your side of the house we're going to leave the app side for the other guys in the software group. But, so DevOps obviously is the mall. You see that with mobile. People who use Amazon, for instance. I have the Node.js, I'm coding this. And the mindset is a developer mindset, iterating. So downtime is simply just reboot, add more code, push new code. You go to an enterprise and you bring that kind of mindset. They want agile, but this idea of the thing could go down is just not an option, right? So we call that no ops, right? So ops dev really becomes the focus. So how do you guys manage that? They want fast app pushing code, but the ops piece has to be reliable. There's SLAs involved, there's maybe some security issues and other things. Can you talk about that trend and how you guys are looking at that? Yeah, it's really, you know, customers are at different stages as you look at that. And you know, there's, you know, the very, you know, the early adopters that, you know, are the Netflix's and what have you. And, you know, they're very far advanced in that regard. And then you look at the other end of the spectrum, oftentimes government agencies, and I've talked to a few here where, you know, again, the walls between dev and ops are still so high and the politics involves and all that kind of stuff. And so it's very, very difficult to push the dev ops a religion now. With that said, they are looking at, you know, improving things like release management, which really is, you know, the, you know, the practice that is being used to, you know, push code from development into operations. And so being able to automate releases, being able to do that for not just, you know, the application itself, but all the underlying infrastructure is a big, big, It's a QA issue on the hardware. Tell them, you know, the networks, right? You mentioned earlier, networks. So doing the release management is one thing, but you got to manage the QA of making sure things are tested. Absolutely. I mean, that's pretty critical. Absolutely. And so, again, that's, you know, that's more on the dev side of the house, but, you know, doing the load testing before you roll things out, making sure that, so, you know, a good example of that is, we showed this coffee app in main stage. So, you know, it said, you know, you guys download this app and then you, you can get free coffee. And what have you had to do, because there's 10,000 people, you know, sitting in front of main stage, everybody potentially, you know, downloading this app, can our servers actually, you know, respond to that, do that demand? So there was some significant load testing involved, you know, given the size of the audience. And so, you know, once that was done, then the app, if you like, was moved into production. Christophe, what's the meaning of business, the term business service management? Is it because you're running the business of IT? I always think of IT service management. Is it because IT is so connected to the business? What's the reason for that? Well, we think that, you know, IT needs to provide business value. And one of the ways you, you know, you can look at that is to say, look, instead of just managing, you know, you know, some infrastructure or the app or whatever, you really want to look at the value that's being delivered to the business. And so, looking at, you know, the service that's being delivered and the value that service provides is what we call, you know, business service management. So what's the, we sometimes, John and I like to sort of talk about the horses on the track. What's the competitive landscape look like in this market? What's going on out there? Help us squint through where HP fits and what's going on. Yeah, so we think we're in a great place right now because there's lots of disruption. And so there's, you know, a lot of movement in the market and, you know, these new trends around, you know, analytics, big data, we believe we have, you know, quite an edge on our competitors right now. And I think the stuff that you've seen on main stage around big data, Haven, and how we take advantage of that in the context of IT management itself is very, very different. Okay, so you've got the big systems companies all play in this space, correct? Presumably. And then you've got some third party guys that play. Some of them are privy partners with you. And then you've got some, I guess, upstarts that are coming in, right? And there's big disruption going on with cloud and big data and that's what you're trying to capitalize on. Absolutely. Good. All right, so what's next for you guys? What should we be watching for? Well, you should watch for, you know, rolling out another round of exciting capabilities around analytics. So we're just getting started. We're just scratching the surface. You know, the visual analytics that we've presented are just the start. We're going to work on that. We're going to work on exciting algorithms that actually over time should allow us to predict where things are going. So, you know, not just drawing some, you know, regression lines, some trend lines, but really figuring out, okay, so if I'm in this state now, you know, what's going to happen in, you know, two, three hours from now? Am I able to predict that there's going to be any issues based on the data I have, you know, similar, you know, similar concepts or setups in the past, you know, can I predict where things are going to go? That's the next step. Okay, Christophe, thanks for coming instead of theCUBE. Obviously, predictive analytics, big data. I mean, this is the future. Rapid development environment, rolling things in production fast. That's what mobile world is showing us. That's what, I'm doing it in reliable. And that's the enterprise value opportunity. Not just the wild, wild west. Some of our folks do it out there. Dave, we can talk about that. Although we love Amazon. Thanks for coming inside theCUBE. This is theCUBE live in Las Vegas. Thank you guys for having me. So as we discover, we're right back on the next guest after this short break.