 Live from London, England, extracting the signal from the noise. It's theCUBE covering Discover 2015. Brought to you by Hewlett Packard Enterprise. Now your host, Dave Vellante. Welcome back to Excel London everybody. This is HPE Discover and this is theCUBE. We go out to the events. We extract the signal from the noise. We're going to talk about cloud. Bobby Patrick is here. He's the CMO of cloud for HP Enterprise. He's joined by Dave McCrory, who's the CTO of Basho database innovator. And Martin Sumner is the tech lead at the National Health Service. Those of you who saw the Olympics, you know, quite an icon in the UK. Gentlemen, welcome to theCUBE. All right, Bobby, you got a great story here. We're talking off camera. Set it up for us and then we'll get into it. Yeah, well, so you know the theme, one of the big themes here, the solution transformation areas. The big one really is transport to a hybrid infrastructure, right? That's going from monolithic, kind of internal traditional IT to one that leverages sort of the right mix of traditional IT and cloud technologies and others as well. And that move is actually one to a distributed system, distributed architecture. One that scales out. One that is always available. One that never has downtime. And one that moves at fast speed, meaning upgrades every week versus every quarter, right? Well, we've been talking a lot about it. We actually have a customer here that's gone through a major transformation of a mission critical system with the UK and HS with Martin here. And he's transformed and they're live and they're in production. They're on HP gear and they're realizing the outcomes. And that transformation story is great. So I think Martin's going to be great there. And then Bashow, their database is one of the key enablers of how that data scales out. And they also, by the way, became a Healy and Reddy certified partner in the last, I don't know, week or so, right? So we're moving fast, how great to have a customer like NHS on. All right, cool. So let's actually go to Martin get the customer perspective. Maybe you could take us back before this implementation. What was life like? Maybe even set up the organization a little bit for those who aren't familiar with it and then paint a picture of your IT organization. Yeah, so the NHS obviously runs all the public health service within the UK. It's one of the world's fifth, five biggest employers I think after the Chinese Army and the Indian Railway. So it's a big old organization. And from an IT perspective, the core of what goes on in real time messaging in the NHS is what's known as the spine system. It was a system that was originally implemented back in 2004. And there's about 20,000 systems across the NHS application instances that talk to the spine. And the spine's responsible for providing a central store of records and documents. So it's a central store of prescription documents, your record of summary care. So what your current medication is, what your adverse reactions you've had in the past is a record of demographic information. And it keeps about one and a half billion records and documents for the 80 million or so people that have been treated by the NHS in England over the past few decades. And in real time, people are coming and getting the information out of that store, updating the information in that store. So that's the kind of spine system worried with them. So it's the nerve center for the electronic medical records. I guess it's unintended, but really that's the central... Yeah, so it's that point of coordination. Yeah, so we know that it keeps everyone coordinated around one individual reference for them and keeps their most up-to-date document and accessible. So when your GP updates something about you, that goes into the summary care record. And then if an ambulance is called for you, they'll access that summary care record and inform the paramedics in transit about any adverse reactions you've had. So at that point of urgent care, they can make that attention. For prescriptions, it coordinates between the GPs and the pharmacies and the payment authorities and this kind of stuff. So it's very much of the hub of everything that goes on electronically within the NHS. And you can hear that we're getting ready for the keynotes. So you can hear behind us. Okay, we go ahead, Meg Whitman again. Martin Fink's going to be on today. So we're going to get this segment in just before the keynotes. Okay, and this is essentially for every participant in the UK. Yeah, and not all applications have been set up to use that central system yet, but there's a gradual rollout going on with more and more systems coming on board over time. And I say that the original project back in 2004, it was a very big, expensive SI program. So one of the largest IT programs in the world at the time. Okay, and so what was the challenge that you were trying to solve and how did cloud fit in? So one of the big challenges that we had is that Spine, its use had taken up and it was now considered to be sort of critical national infrastructure. It's really important for the country that Spine keeps running, but it's costing a lot of money. It developed on sort of 2004 tech stack. We've got very expensive infrastructure and we need a lot of people to be around, to make sure that they're in case the things failed to get it back up again. So we wanted to rebuild that. We wanted to reduce the cost of running it and in particular, reduce the cost of running it at that high availability. So we want to maintain that availability, maintain the protection from data loss. That's absolutely critical to us, but we want to be able to run that at lower cost. And so we started a project internally to say, okay, rather than starting with the kind of enterprise platforms that we had from 10 years ago, why don't we look out at what the big internet companies are doing, the sort of patterns that they're adopting, not just in terms of the technology that they were using, but their approach to releasing, their approach to development and that kind of stuff. Why don't we go and learn those lessons and see if we can build a better Spine. So we started with a very small team just with two people originally and we built up from there and started proving as we went that we could build a better Spine by adopting these patterns. So there were DevOps Ninjas, two people. Yeah, yeah, yeah, we started off with our operational tool and thinking about how we're going to do continuous delivery and this kind of stuff and build that in from the start. And by the time we were live, we were up to 35 people. So still a fairly decent size team, but that compares probably by, I think in the original contract at one stage there were 2,000 people working on it. So it's a big difference in the volume your people require to build and manage the service by adopting this change of pattern. And Dave Basho provides the database for this new infrastructure. Talk about where you fit in this whole thing. Sure, so we're the underlying data store for all of this. There's a lot of effort that's been put in by the NHS to really do all of the things they could on top of the distributed database that we built, which gave them a foundation for the redundancy and scale in the storage that they needed on top of HP Gear. Okay, I got to ask you, why Basho? Why not one of the other zillion database companies or Oracle or? Yeah, I think what we really liked about it was it's focused on the operational aspects. Yeah, for government systems, when we build them, we expect them to last a long time. A lot of enterprise IT, you rotate it quite quickly and this kind of stuff. And so all the cost is going to be in that long tail of how long we spend in operations. And when we first started working with it, what really appealed to us was, and also the themes that we kept coming about hearing from customers online of Basho, so other people that were using it in the mobile app space and this kind of stuff, they kept saying, well, a node failed last night and we didn't care. We didn't notice until a day later and this kind of stuff. We got to sleep last night. Yeah, we wanted a system where we wanted to be highly available. We don't want to have a load of people around who have to care immediately if something fails. And those are the characters that gave us, and we really liked the simplicity of the solution. We felt it was a really simple tech platform that did just what we needed it to do and not an awful lot more. And at the time, in the early versions of React, especially that's what it did for us, so that was really appealing. And so we first started experimenting with version zero dot 14 of React, so it was quite early days in that stage. And I think a lot of the other key value stores were likewise very early on in their stuff. And we just really liked working with it from the off and we really felt that it was going to support us in operations in the way that we wanted. Okay, and you would consider this a private cloud or is there sort of a hybrid nature to it? A hybrid infrastructure is the right way to word it, right? I think that's the way the word it is. It's a multi-datastere implementation, right? So they've got HP, you know, relying gears in different locations. You need an application, a distributed software stack, though, to take advantage of that in an internet kind of, an agile kind of way. So you have to have a distributed database and you've got to be able to fail over between centers so you can do upgrades in certain centers. And that's really what the story of hybrid infrastructure is all about. The software's got to be built and the developers and the ops teams have got to be able to interact with that technology in a way that's new-style. And this is what we're talking about to discover the entire conference. This is what it's all about. They've just delivered on it. And the stats, some of the stats that Martin has, like I don't want to quote, but like the stat of 60-minute queries go into two seconds, right? Yeah, how'd you measure the business impact? So, I mean, one of the primary things is in terms of the difference in performance, that wasn't a target of ours because primarily we wanted availability, but in terms of difference in performance, we've seen a 90% reduction in response time across the board. Now, when you total up the number of interactions across the NHS, basically the NHS spends 800 working days less each day waiting for a response from spine. Each day? Yeah. So, you know, there are lots of small, small differences that build up over time. That's a business case right there. Yeah. That's a telephone number. So, to make that, that's made a big difference, but none of that would have been worthwhile if we didn't have availability, you know? And maybe we could have chosen another technology, maybe done it even faster, but the main thing is we've got to maintain that availability. You know, it's got to be there 24-7. You know, you don't want to be in a situation where you're on the ambulance and the system wasn't there to respond and tell them your information. So, it's a big productivity hit in little pieces each day, and you've also dealt with the IT labor issue that you had. Yeah. So, how did you marshal all those extra resources? I mean, it's the UK, I'm guessing you didn't lay off a zillion people. How did you redeploy those people? No, I mean, I guess, you know, because it's so dispersed, you know, we've got about 200,000 users. So, we're just making 200,000 users' lives just that slightly bit better, you know? And hopefully they can then spend more time, you know, if they're a nurse or a GP or that kind of stuff, a little bit of extra time with patient care, which is really what we want. So, if we can shave a few seconds of the time the GP spends entering the patient's details into his computer, he can spend those extra few seconds with the patient, then that accumulates for us. So, we don't necessarily want to make the saving of cash, we want to free up time with patients. So, that's really the outcome that you're looking for. Yeah, saving lives. The quality of patient care, saving lives, responding more quickly to requests. Yeah. Okay, and now, when did you go live? So, we went live in August of last year. So, we've been live now for about 15 months and we've been effectively fully available since we've been live. So, what's the feedback been from the user community? We don't get that much direct feedback from the user community. It's mainly the suppliers that integrate into it and obviously they've seen a big improvement in performance but kind of the best feedback we've got is no feedback. It's no feedback. Yeah, that's exactly the type of feedback we want. It's one of the things whereby it could only, you know, if it went wrong, we're in the papers. Yeah, so, if we're out of the papers that's great feedback for us. You know, it's available and it works and that's the important thing. All right, so let's take it to HPE Discover. We're getting ready for the keynotes. Bobby, what have you seen, you know, on the floor in customer meetings? What are people talking about? Obviously the split, but what else? No, they want to know how to go to a hybrid infrastructure. And I think that's one of the reasons why Martin is so important because it's not just, you know, by some infrastructure. It is learn how to program, train your developers, the culture impacts. We've got a bunch of DevOps tracks and booths this time. They're packed this time. You know, a year ago, they weren't packed, right? So I think people really are buying in and they want that speed and agility that Martin talked, but they want to be able to upgrade a system and not have downtime, right? And so now it's really how do I get there? And you know, on the keynote yesterday, Antonio Neri talked about, you know, every company has their right mix of hybrid infrastructure. So it's very important to us to help customers understand what is your mix? Based on what systems you're running, what traditional systems do you have, like SAP? You know, what kind of cloud-native skills do you have? You know, I think it's important for us to understand every company's mix is different, but we do want to bring these best, these proven best practices and figure out how to make them replicated, how to show people a path where they can actually realize the benefits that we're talking about on stage and it's not just, it's not pixie dust, right? It's real and companies are realizing it. So I think people are moving into this phase of I want, I want, I buy in now, how do I get there? And Dave, recent HP partner, Helion partner, you've been a partner for a while, but Helion certified. What gives you confidence as a HP partner that you're going in the right direction with HP? HPE, sorry. Ladies and gentlemen, please take your lead to the general session. We got a couple of minutes, we're good. So you can hear back around. Should be great general sessions this afternoon. Martin Fink, I'm sure is gonna be teasing us with the machine. So I think the biggest thing is the fact that HP gets the whole approach that people are moving to as far as this cloud, this hybrid approach to things, trying to operationalize and go for efficiencies. And that's really, that's kind of where the rubber meets the road. And I think that gets lost in a lot of the messaging in IT and elsewhere. It's, oh, we're gonna build this new shiny, awesome thing. But how are you gonna operate it? How expensive is it going to cost? What's the management going to be like? And putting something like React, our product on HP just makes a lot of sense. Kind of moving to something that's really a gold standard of what you would want to operate on. All right, and Martin, it's less than 30 seconds, but what do you want to see from HP? Did it help you solve this problem? What's next? And I guess, yeah, we're always looking at a couple of things. Firstly, as Bobby said, a big focus to us is being able to change frequently and at low cost. And that was always a blocker in the past. The other thing is that we're in a constant battle on the security front. So we've always got to try and look to improve security. And I think that's one thing we're going to talk to HP about how they can assist us more. That's one of the rooms here. We'll check it out. All right, Martin, Dave, and Bobby, thanks very much for coming on theCUBE. It's great to have you guys. It's great to be here. All right, keep right there, buddy. We'll be back for the keynotes right after this short work.