 Live, from Las Vegas, it's theCUBE, covering NetApp Insight 2017, brought to you by NetApps. Welcome back everyone, live here in Las Vegas, this is theCUBE's exclusive coverage of NetApp Insight 2017 here at the Mandalay Bay in Las Vegas. I'm John Furrier, the co-founder of SiliconANGLE Media and co-host of theCUBE. My co-host is because Keith Townsend at CTO Advisor and our next guests are Josh Atwell who's a developer advocate at NetApp and Jason Benediches who's with Principal Consultant, A&S Group Cloud Service Provider in the UK. Great topic, talking DevOps. Guys, welcome to theCUBE, good to see you again. Good to see you as well. Thank you. Boy, DevOps is gone mainstream. Yeah, it's a thing. It's absolutely gone mainstream. We've been saying it for years, remember going back a few years ago, you say DevOps, huh? Infrastructure is code, everyone loves it. That's now the new model, people are moving fast too. What's going on with NetApp and tell us your story, good? So, within NetApp, we look at DevOps as a unique opportunity for us to level up everybody that's doing infrastructure and going from saying, you're just going out and developing an application to saying, we can actually help deliver you the best experience. When we look at where applications are being developed and supported, everybody likes to say it's straight up to the public cloud. That's where all the innovation happens but it's also happening on premises as well. The reason that we see most frequently is that reduced friction, going to the public cloud, that has become a model that people can go out, they can get what they need and do what they need and it's been something that's significantly easier than what their local IT organizations has had. DevOps is forcing infrastructure in IT to understand that availability and reliability, which is what we've always been measured on is no longer the core measurement that we have to focus on. It's agility and availability and delivering unique services. Well, I would just say to your point, Wikibon analysts research have validated your point and they actually show in the data that the on-premise, they call it true private cloud numbers, are growing actually, not declining. What is declining is about $1.5 billion in non-differentiated labor but that's shifting to SaaS models. So what it means is the on-premise action in a cloud operational way is growing, which is not saying that's declining, it's just saying people are getting their house in order. They're doing DevOps on-prem, prep to do cloud. Now the cloud's got native stuff, you do bursting, you can put some stuff in the cloud test dev, sure there's great use cases but most enterprises are on-prem, getting ready to take advantage of it. It's an absolute and conversation and that's also something that we are working really hard with our customers and our field and the entire company as a whole to understand it's not an or conversation. Most companies are looking at how do we solve a variety of different challenges, how do we accommodate for a variety of different workloads that are being developed, and how do we modernize the mode one operational workloads that we've had and bring them into the future with new services. So it's an absolute and conversation. It's a pretty exciting time to be dealing with IT. So Jason, as we think about DevOps, we have plenty of examples for private cloud and inside of our own data centers but you help run a public cloud. So we help run services within a public cloud and a hybrid model. So we run a number of services to man assessments. So we help in the UK, I think we're probably a little bit further behind than the US is currently. So some of the biggest services we do is helping people to assess their applications, assess their data, understand what they can move using things like the gardener time analysis where we can take best leverage of on-premises private cloud, where you've got hybrid approach, where you've got native. So we got the expertise around retooling and assessment services to move legacy applications into a cloud model and then we provide management services on top and those sorts of things. So that's where we utilize the DevOps around taking what would be our managed services ITIL processes, the things that people would traditionally do manually. We take a lot of that and we pre-package that up into workflows and data automation operations for our customers so they can provision where they like across a multitude of on-premises and in the public cloud. So we take that work that would traditionally be done by an analyst on a desk or that sort of thing, package that up using a lot of NetApps APIs and the Solifier tooling. So we're saving enterprises time so they can work on what's really important to them and that's their line of business applications. From an assessment perspective, I'd love to get feedback. What are customers learning? Like is it that they thought that they could just lift and shift or that they have to go through some type of DevOps transformation? What's been the balance of the results? Yeah, so a lot of people don't necessarily understand where they are. There are a lot of misconceptions around being able to lift and shift things to the cloud but that's not really a great cost model. So, and also I find in the public sector in the UK a lot is you've got a lot of legacy applications that potentially people don't have any knowledge of because the people that ran them and installed them in the first place have long gone. So they need to understand what those applications do for their business, what the business processes around them are and how they can take that forward into a new model. A lot of retooling, actually a lot of the time we see that that application should probably ditch and let's look for something that we can just build cloud native. That requires a new set of skills to operate at the higher level of the stack as we call it in the industry. However, that leaves a lot of low level work that still needs to be done. So automation has kind of walked hand in hand with DevOps. What is the NetApp story around automation and helping to remediate some of this low level activity that needs to be done repeatedly? Big focus for us as a company is not trying to dictate tooling to people, right? If you are using Docker, we offer a native Docker volume plugin that allows you to plug right into Docker and be able to provision and manage storage as an application owner or developer to get what you need and to handle the services that are available there. When we look at configuration management or helping code and artifact management, cloud with OpenStack or VMware be realized sweet. Our initiative is to make the NetApp products seamless and invisible into your processes. How do we remove and eliminate handoffs and how do we make all of those processes effortless so that as you identify those tasks and those high effort but low value tasks that has to be taken advantage of. And automation's critical there. Yeah, yeah, being able to automate those things, remove people from that process and using their skills and talents for things like auditing and understanding proper behavior, checking that people are delivering what they are supposed to and consuming from a policy framework. I'd like to get back to the automation but I just want to shift to Josh to hold the thought on automation. Josh, I want to get your thoughts on, because we get the automation, we start talking about hybrid cloud, right? So you're doing hybrid cloud. You're on the front line, you're doing it. And also hybrid cloud also means things differently. So when you think about hybrid cloud, the customer's got to get their act together. So we heard earlier from the NetApp folks, the VP of Systems Engineering, we're doing three things, modernizing the infrastructure. That's just like, okay, go clean house, fix things, making sure we're solid, rock solid, build the next generation data center, be ready for the cloud. Yeah, okay. So there's some things that need to get done there. So what's your view on the table stakes to get there? Because you got orchestration capabilities, cloud orchestration, demo is hot, you saw that at the show here. What is NetApp doing to make hybrid cloud easier? So across all of the products that we utilize wrong with NetApp, you've got APIs and everything. You know, they've got a lot of really good tools there. And they're moving away from the traditional hardware. I've been working with NetApp for like 16 years. So, you know, it was a hardware company, a software company, and now it's just moved on even further. There's a further evolution there of, you know, a management company. It's not just, you're managing your data, the data flow, the fabric around it. And the tools that are on offer there are just game changers, especially the cloud automation option this morning. Yeah. I mean, as people know NetApp, I mean, eight years ago I was scratching my head saying, wait a minute, why are you going to Amazon? So early in cloud, so clearly they know what DevOps is. So it's not just lip service. We know that. That's just my personal observation and experience with NetApp. But Josh, I want you to talk to the audience that is either a NetApp customer or looking at NetApp, what's different now? I mean, what should they know about the new NetApp now? Obviously, you're on the A team there, see the shirt there, but NetApp has changed and they're changing. I mean, SolidFire came in, you're seeing a lot more action on the DevOps cloud, the Flash and good stuff there, but NetApp has been an innovative company. What's the new story for NetApp in your words? For me, it's the speed that they're able to react to the market. You know, removing the on-tact recadence model, you know, six month releases, moving products away from tin into software. It's all about the value of what we can provide. You know, we've got standalone products now from NetApp that can just do Office 365 backup. You know, that's something that's completely moved forward. You've got a level of innovation and speed coming out of NetApp that's just unrivaled. I mean, Josh, look at your thoughts back to the automation now. I'm a CXO, the constant thing I hear all the time is the following narrative. I don't want to show any new toy, I got a lot of stuff on my plate. I got an application development team I need to scale up and make modern, which is DevOps. Not just take the old guys and put them in, I got to recruit, retrain, replatform. I have cybersecurity going, I got to unbolt that from IT and make that essentially a top line, top reporting to the board, do all the cyber stuff. I got to get data governance stuff to deal with. And by the way, I got IoT over the top coming in. I don't, if it's not clear as day on the cloud, it doesn't meet my conversation. How do you guys engage in a dialogue like, why do you agree with that? That makes that statement. But that's a lot of stuff going on. Bombs are dropping inside the customer's environment. They're like, this is hell right now. I got a lot of stuff to do. How do you guys help that environment? But I think one thing that we have to be mindful of is that we've moved beyond being able to define a very static and rigid infrastructure architecture. In the past we would define what our storage, what our compute, what our networking is, and that's what it's going to be. It's very easy to say, I know how to support 10,000 exchange users. That's always been something that we've been comfortable talking about. What you outline is the new reality for IT in that we are getting a diverse set of requirements where we'll come in and say, we need to deliver this new application so that we can get to market and capture this while I was actually talking to someone with the military. So what if the military was to develop a new recruiting tool? And they go in and say, we need to build this recruiting tool, but we actually don't know how much data is going to be required for it. IT is not comfortable with that conversation. But NetApp has developed our portfolio and the integrations and tool sets that we've integrated with to make that conversation a little bit easier. They're not comfortable because they can't forecast it or it's a blank check in their mind or they don't know how to architect it. What's the... It's because we're not accustomed to architecting for those types of scenarios. We generally have focused on what is going to be your use case? When do you need it delivered by? How much do you need? We're still having that same conversation but the answer now is, I don't know but we have to be ready for whichever direction it goes. That's a good point. At VMworld, we notice that there's a convergence and not a lot of people are talking about this yet but I can see the canary, the coal mine chirping away is that the convergence between hardware and software stacks are coming together. There are untested use cases coming down the pike. I need this but we haven't tested it or we don't know the capacity. So you have a serverless mindset, you got to have DevOps mindset, you really got to be prepared. Well, there's certainly a lot of maturity that we're working through. We are definitely from a DevOps perspective in that juvenile phase where we're learning who we are, the changes that are happening to us as we go and we're getting a much more responsible view of what we're trying to deliver against. It's really uncomfortable for a lot of people to have a conversation where there's so many unknowns but fortunately the technologies we're able to bring to market and deliver are providing, as I describe it, a foothold to make you feel stable in that process or at least know that your data's getting where it needs to be and protected. I know, we see that evolve in the customer mindset too where you start to see the word trusted relationship become real. It became a cliche, oh we're a trusted partner. The reality now with all this uncertainty they need the headroom, they got to cross the bridge with the future with proven people. So that's why I kind of like, I don't mean to dis on the start out but the shining new choice not going to win the day. You got to really hit the scenario today and prepare across that bridge to the future with partners, I think that's what you're saying. Yeah, that is a big part in the partnerships that we have with folks like Red Hat and JFrog where we're trying to improve that experience of implementing these environments and supporting these new workloads is absolutely a big part of what we're doing. So I'd like to talk a little bit about the necessity of requirements coming from the business and tying it into something I heard from the stage yesterday, I'm not a storage guy. I'm a data guy and you've said that before but one of the things that has interested me is this concept of the data fabric. Can you tie in the vision of data fabric to kind of this model of DevOps and being able to adjust to the changing needs of the business? I think what's really important to be mindful of is that as we are seeing IT getting these requirements, as the businesses are identifying what is really impactful and the innovation that we need to deliver on, the data fabric is providing choice. It's allowing you to look at being able to deliver these enterprise class protection and replication and capabilities and allowing you to develop, innovate and run your workloads wherever is most important to you without having to completely re-shift your thinking and what your skill sets are. And we are able to level up everyone that has been involved with NetApp and has invested their career and invested their energy and becoming knowledgeable in that space. Now allowing them to extend out into new areas in the cloud, you know, hybrid cloud frameworks, but also providing these capabilities to the people consuming those resources without them having to care about the infrastructure. They know it is there. They know they can reach out to it and, you know, define snapshotting and take advantage of clones and deliver a good developer experience without having to, you know, understand exactly what's happening in the infrastructure. Thanks so much for coming on. I'll see you having seamless infrastructure is what everyone wants, but it's hard. Yeah, final comments as you go into the future now with DevOps, I mean, it's become operationalized. A lot more work to do. It's not that easy. What's the hardest thing about DevOps? Final comment, you guys each weigh in and get the last word. What's the hardest thing about DevOps that people may not understand? Because it sounds so easy. It's magic. I think the hardest thing for most people is having a critical eye and being pragmatic about where the challenges really are. If you look at the methodologies that DevOps promotes, it is really identifying the constraints in the workflow process. So regardless of what you're developing and what you're doing, being very pragmatic and realistic about where those constraints are and focusing energy on solving for those constraints. I think with what we deliver out to market, we are providing people some stability so that as they are going through this process and things feel really shaky as they accelerate their pace of development and release of software, they have some stability so when they focus, they don't feel like the wheels are coming off the car, if you will. I think what I find is that people are all they need to understand, DevOps isn't something you can buy. You need to build. You need to get the right people. You need to get the right processes, the right mindset and embrace it. A lot of people think it's just, you see job adverts these days. I want a full stack DevOps engineer and it's just not that simple. You've got to take the time, take the effort and move with it and learn as much as you can. And it's a talent issue too and I'll guess one final, final question because it's just popped in my head. At Big Data NYC last week in New York, what became very clear to us was, certainly in Big Data applications analytics, a lot of things are being automated. But question for you is, when should you automate? One comment on Big Data NYC, a guy said, if you do it more than twice manually, automate it. Not that easy in storage and networks and data, most DevOps guys have an eye for automate that. Like they see it, they automate it. What are some of the things you see being automated away? Is there like a ethos? Is there like a saying, you automate twice? What's your thoughts on automation? What should you automate? What's the order of operations? What do you go with? What's the low hanging fruit? With respect to DevOps in particular, it is truly finding the constraint. Identifying areas where people are becoming a bottleneck in processes or the process itself is a bottleneck to success. Focus on that area first. Now, it's also easy to just try to pick the low hanging fruit and do various things, but there needs to be a discipline in looking at where are your actual bottlenecks and how can I remove those bottlenecks? So you can't read in a blog post, you got to know your environment, see the pressure point constraints. Yeah. I mean, you can get some direction advice, but you're saying, look at your environment. Yeah, we're now moving away from a world where virtualization allowed us to just throw everything into a big resource pool and we just didn't pay attention to it any longer. We are now actually having conversations again. Yes. It's engineering involved. Not just writing some code. Josh thoughts on automation? What you automate first? I share a lot of those things. You need to look at your processes. You need to look at where you've got your bottlenecks, like he said. You know, things that we would traditionally do in the past as a service provider where you've got teams of analysts and engineers working on things. If you can speed that up and allow them to provide a better service to your customers, then yeah, well, certainly work on that automation. Deploying out new models, new even internal stuff that we need to deploy out. If you can, if you need to do that more than once or twice, you know, for test environments or those sorts of things, then yeah, certainly all made out because the more time you get out of your people, the more value you're delivering to the business. Thanks, Josh. A team, one of the quick sound by what's the A team? Is there a certification? Is there a bar to get over? There is a pretty high bar. It's an advocacy program. It's quite a small, tight-knit group of partners and customers of NetApp, and we work in a 360 feedback loop between the NetApp product management teams and other developers, and just give feedback and then rave about them when we feel it's necessary. Have a beer or a coffee and tea and say, love when a plant comes together. Yep. Good resistance. And so John had also mentioned, NetApp has also delivered a developer in an open-source community called the Pub. So NetApp.io, it's a location. We actually have the code on bar behind me where we've got people that are coming in who have interest in containers, interest in OpenStack, DevOps, and these new models. We have a large community, over 900 people participating. The Pub. The Pub. Is that right? Yep, NetApp.io. NetApp.io. And just, you know we're data-driven, you've been monitoring the community's data just anecdotally. The favorite drinks of developers in our community, beer and tea. Makes sense. Pretty makes sense. Beer, obviously. Tea over coffee. Slow release caffeine, I think that's probably what's going on. Thanks guys so much, Josh and Jason. Data from the field, from the front lines on cutting-edge DevOps is going mainstream. This is the cloud-native, native cloud on-premise infrastructure innovation here at NetApp. I'm John Furrier, Keith Townsend. We'll be back with more after this short break.