 Live from Las Vegas, it's theCUBE, covering NetApp Insight 2017, brought to you by NetApps. Hello everyone, welcome back to our live exclusive coverage of NetApp Insight 2017. This is theCUBE, I'm John Furrier, the co-host of theCUBE, and co-founder of SiliconANGLE Media of my co-host, Keith Townsend, CTO advisor. Our next two guests is Dave Hitz, who's the co-founder of NetApp, and Anthony Lai, who's the EBP and cloud business unit manager. Welcome to theCUBE and welcome back, good to see you Dave. Thank you. I always love, I wrote a post years ago called keep the founders around, you know I always joke with you on this, but the DNA of a company is super critical, and how the products get positioned even as the evolution, the DNA is critical, great to see you out on the front lines, pressing the flesh with the customers here. Keep the founders around, so I have a theory about that, because some people say companies where the founder stays around are more successful, and therefore I must be awesome. I have a different theory, which is companies that are really successful are a more interesting place for founders to continue to be interested to stay. So I think that the causality may be the other way around. It's like if your founders want to keep staying and playing, you must be doing really cool stuff. It's a cultural issue, and this is a big DNA discussion. We go back seven years, we've talked, and I've talked with your former CEO, Tom George, who's about this. Why are you going with Amazon? Everyone's saying that's a bad move, but if you're a contrarian move, you guys said, hey, the customers are asking for it. Now it's all cloud all the time, data as a fabric. This is now mainstream. Really good tailwinds for NetApp right now, because you've got the core base, the Shenian new toys not winning the day, but blocking and tackling good technology and the right customer focus. Talk about the cloud impact, Anthony. Yeah, I mean, just to make a point, just on the last comment, I mean, what Dave does, I think is you lean into things that are disruptive, and I think very few founders have that ability to sort of... Well, sometimes I think the biggest value add I can bring to NetApp is to give people permission to let go of the old stuff, and some of it's hard. I mean, I'm the guy that wrote Waffle for ONTAP, and so I'm not saying, I mean, we're still shipping a lot of that stuff, and it's awesome, but some people struggle to say, what do you mean we're going to sell another storage system? This is always the best one for everything. That's what we've been saying for so long, and so if I can let go, it's like it's my baby, and I still love it, but can we have another kid too? I mean, you've been instrumental in the cloud strategy, and you tell a cloud story first, and it's not what you would expect, and I think that's what gives NetApp, it's sort of unique, and I think it's 25 years, is you go out and you could easily talk about all the things that NetApp has done, but you choose to talk about where you think NetApp has to go. What was interesting to me about today's general session, because we had so much new stuff, I think you almost can't get your head around it, we had to divide it into categories, and the categories we chose really align with how we see customers working, and so the first category is a lot of people have, and will continue to have for years, the traditional style of data center with client servers, Linux, Windows, you know, you rack and design it, like what should the fiber channel be, it's virtualized, but here's the chunk for Oracle, here's the chunk for virtual desktop. It's running apps, by the way. It's running critical apps for the company. All of this stuff, and then you've got this new style, which is all one you racks and wired to the top HCI, and this whole next generation data center, and then all the cloud stuff, that it's services running entirely in Amazon, we've got services where we're moving data from one hyperscaler public cloud to a different hyperscaler public cloud, with no net of hardware involved, I mean, these are entirely cloud native cloud resident services. I want to solve like one from like one region to another region of AWS, so you're saying that the solution can move from one cloud provider to another. We've been doing that for a while, I mean, ONTAP itself, you can buy ONTAP cloud for AWS and you can buy it for Azure, and so you can establish a cluster on one and connect it to a cluster on a different one, and let ONTAP snap between the two, move workloads between the two, back up between the two. We've always had that now, the orchestrator that we show today pushes us much, much higher and provides our customers with a true multi-cloud platform, but a multi-cloud platform that really starts to blend compute and storage together, and it's a platform that's built from the ground up on Kubernetes, which is now, I think, the sort of universally accepted container strategy for microservice-based applications, and yes, that platform will allow you to deploy an application package at the same time on any of the big three hyperscalers. So a lot of the pushback that I saw on social media was from the announcement yesterday with Microsoft Azure NFS. Why are, why is that? You got pushback? Yeah, pushback. We're like, well, the object storage is the future and it's the best way to do cloud, period. Actually, it was the only way. Can you talk about the importance of NFS and the data fabric? Well, can I back up a step? Just to be clear, object storage is awesome. It is. And NetApp has an object storage solution and I'm not going to diss object storage, right? It's great. However, NFS is cool too, and a lot of people have a whole bunch of apps on-prem and they've written them already. They run whatever they run, and if it uses NFS and you'd like to have it in the cloud, you don't want step number one is let's rewrite it. Exactly. You want step number one is it already works and I would just like it to be working over there so I don't have to mess with physical hardware. I know this might be sacrilegious for me to say it's been from Silicon Valley and UR2 but the shiny new toy doesn't win the day and what we learned from the Hadoop and we've seen it a little bit in OpenStack but they caught it early before it became a tumor was the cost of ownership to write stuff from scratch is problematic. So there's an issue of legacy's not a bad thing. Look what containers, your point about Kubernetes. So you have to run these apps. No one wants to run my code and argue if it's a bad thing or not a bad thing. It exists. Exactly right. And we want to help take care of it. But why rewrite code is a mandate to get this? So nobody, I mean if it makes total sense, okay you look at it, but it's not. No I think IDC, IDC pegs, file based workloads at more than 24 exabytes with on-prem growing at somewhere around 18% Kager and cloud growing at 25%. So objects are not the answer to everything. Old or new actually. Is it an application developer? I like the opportunity to have both and I think applications will consume both. Let me jump into the announcements that are on stage here, the other conversations. There's a lot of stuff as you mentioned. So the folks should look at the keynote, go to this, we've streamed it live so you can go to SiliconANGLE or go to netapps.com check it out. But a couple things jumped out at me. The on tap nine point was a nine point three and solid fire, interesting integration there, shipped, great stuff. The cloud orchestrator, seamless moving data across multiple clouds. Everyone knows me, I've been critical of this. This is, I've been looking for someone to actually show me, because multi-cloud is hard. You have latency issues, there's a ton of stuff but you're not rewriting code to do it. You can do it on-prem, huge deal. And then the other thing is just the general sentiment of the 18 guys around the channels. The channel partners are energized. They see an opportunity to build a business sales channel for netapp, but more importantly, they can come in and deliver to customers. Guys, unpack that, those dynamics. Obviously the solid fire thing of flash. Can I start with the channel? Yeah. When I look at how the channel interacts with a lot of customers, they make their money selling stuff, often gear, but if you look at what are they really providing? A lot of them are acting as IT consultants in some cases with smaller companies as CIOs for hire. And so, you know, it doesn't, people go, oh well what do they do if it's cloud? What do they do if it's on-prem? It's like, the customer still needs that same advice and consulting. CIO has cloud concierge. They have their own cloud service for their customers. And so I just think there's a big opportunity for the people who choose to embrace it. Anyone who's telling their customers, whoa, whoa, whoa, slow down. You don't want to go in the cloud. We'll help you not go in the cloud. Like, I don't think that's a long-term business model anymore. Cloud is destination, it's happening. There's a thing I would say on the partner side that we've seen is that we now have, I think, credibility in the cloud. So much so that we are signing partners that only work in the cloud. So a lot of Amazon partners, a lot of Azure partners have come to us and said, hey, you know, we didn't realize you had all of these data services and we are, we're running customers' infrastructures on the hyperscalers and we'd like to use your software to make our lives easier. We'd like to use Ontact Cloud, we'd like to use clouding. So as well as our traditional partners, there are other partners here at this event that are first-time as an insight. So talk about the cloud dynamic because certainly it's a lift. Rising tide floats all boats, a tailwind, whatever you want to call it. But now I'm a CXO, I'm having a conversation. Like, whoa, you got my attention, NetApp, all my old trusted NetApp guys, the storage guys, and they're talking data, music to my ears, because I got all this stuff going on, GPPR, and all of a sudden cloud, I didn't know that, I didn't know they had a cloud. You don't get a cloud strategy. You either do cloud or you don't. So this has come up on theCUBE a lot. Talk about the dynamic, how you talk about the dynamics. I'm like, okay, I know I got to build for the cloud. How does NetApp fit into my strategy? Because I got to cross the bridge to the future. I got business to take care of today, both on-prem and the three pillars, but I got to have a cloud vision. Well, let me back up a little bit. One of the reasons we think we can help that we're very well positioned to help, it's very easy to fire up 1,000 CPUs in the cloud. You want 1,000 CPUs, you fire them up, then you unfire them up and everything is easy. Until there's any data, what did they want to look at? How do you get it in there? What did they create? How are you going to keep it safe? Do you want to leave it in that cloud or a different cloud? Or do you want it on-prem or all three? And as soon as you start getting yourself into those questions, you go, whoa, that's the hard part of the cloud. The good news is, that's exactly what NetApp does. That's the kind of work that NetApp focuses on. And so the starting point is, look, the hard CPUs, computes, lambdas, a container, all that stuff is easy until you get to the data which lives forever and you're legally required to do something with it. Now let's talk about what you're trying to accomplish and where you're going, like that now is, and one of my goals these days, how long can we talk without mentioning a product? Because it's not, you know, eventually you're going to have to get to, oh, by the way, we have a backup tool that'll reach into Office 365 and suck it out as objects and put it on your on-prem object storage. Well, backup is a whole other story. It's AWS or something like that. You know, walls in a cloud. So eventually you get to some tool or some product but you want to talk for a long time about where they're going, what they're trying to solve, what they care about. Often they don't care about a thing you think they should. Like, aren't you really concerned about budget? No, actually we're dying because we can't solve this problem. So the budget comes after we solve that. Okay. We were talking last week about, I was calling it the toolshed paradigm or paradox. And the toolshed paradox is that they're focusing so much on the tools that they have this bloated tool chest. Somebody's getting collecting dust. They bought a hammer that they're trying to mow their lawn with. So like, you have a problem of too many tools. And pun intended. So the question is, is that as it kind of distracts from the focus to your point, data. Data seems to be the killer app in the cloud because now, not just moving data around cloud, developers are using data in real time. So you batch in real time or it's huge. How is the application developed? Because I'm a CIO. I've got a lot of things going on on my plate. I'm ramping up DevOps and more application development, new developers, open source, blah, blah, blah. Security, governance. Well, I mean, to me, I sort of think a really nice soundbite that I got was, I was an application developer and my career has always been building applications and it's always been the applications that own the data. There was an application server and it executed business logic that read or wrote into a repository. A data. I am at the point where I believe we are in an inflection where now the data will own the application. And what I mean by that is the data has to be fluid and available for many applications to consume it. Some of them will enrich it, some of them will replace pieces of it and so architectures have to change and I think NetApp's incredibly fortunate that we have such a strong data story at a time where the data itself will be the primary asset on a company's balance sheet. I really believe that. If you believe that, if you believe that point, which I do by the way, I think you're 100% right, that changes the paradigm, flips it upside down. But it also creates the conundrum of data governance because I got a policy, I'm going to put the brakes on that because you're freeing the data to be addressable, to be more alchemist kind of model where I can't control it but I need to control it because I got regulations, I got governance issues. So, should you pause? How do you guys address that? I know you got a governance story but that's a dynamic, it's a psychology. Well I think, yeah. But add on to that, talk about availability and governance. So there's the policy piece of it and then there's the availability piece of it just because I can move from an application developer's perspective just because I can move an application to the cloud doesn't mean that it will perform like it will when I, you know, using 100 microseconds of latency in my private data center. So how do I get the policy and the technology governance that combine together in the cloud? I mean, I think I'll make two points. I think the obvious answer to the first question is we have the data fabric. And I think NetApp has pioneered, you know, its strategy around a set of data services that do certain tasks that can be consumed as applications or as APIs. But then we've gone one level higher and now we orchestrate and connect those things up and provide meaningful solutions. And Dave has a fantastic, you know, we were talking about a fantastic demo with Storage Grid, I'll let Dave explain that. The second point I would make though is what you've got to understand is that the customer that we talk to isn't AT&T, that's just a big building with a logo on it. A customer is the person inside the organization and we all now know that there is a new, there's a new customer and that customer is people refer to as the data scientist. And there haven't been data scientists before but now every company is hiring data scientists. Why? Because the data itself has become the primary asset. Application developers are now serving the data scientists. So DevOps was developers making infrastructure as code with operations. You're essentially describing a new paradigm, data ops. Correct. Data as code. Because you need to have it programmable. Absolutely right. And I think that's what most people call metadata or they talk now about APIs for everything. And so I think that's the new norm. I think that there will be very large, you know, catalogs of data surrounded by policy and government governance but expressed essentially as an API. And that the data itself can be manipulated in real time or through batch using a set of restful APIs. And I think Dave, you know, you should share, you know, the demo, the storage grid guys today. It's just a fantastic data fabric use case. Some of my favorite use cases with the data fabric is where you're confused. The line is blurred even. Is it cloud or is it on-prem or what is it? And we've been working hard to integrate those things. So here's an example. We showed, and this is a made-up use case, but it was an on-prem solid storage grid. So it's a bucket of objects. Did I mention we love objects? It's a bucket of objects and their faces. And the problem was how do we identify what's going on with these faces? Are they happy? Are they sad? Are they angry? And you don't want to write your own face recognizer and Amazon has good face recognition technology. Recognize. And so the use case that we constructed is, here's the bucket. We have integrated our storage grid object storage with Amazon's simple notification service. And so anytime a new object gets put into the bucket, it notifies Amazon. Amazon can do whatever it want with that information. Hey, here's the bucket, here's the new object added. What we had it do was issue a lambda, connect up the notification to a lambda, have the lambda come back out, grab the data from on-prem, look at it with the face recognizer. Okay, happy, and then go back on-prem and update that metadata. So is that cloud or is that on-prem? We used Amazon's lambda. You're hitting the new development. This is data fabric. This is the new development reinvention. This is what I think a renaissance is coming big time. Because making that happen, it takes creativity. The barriers to pull that off now are almost down to just knowing what's available. And so I think a renaissance is coming because that's amazing. And then you got to say, how do you scale that? And this is the challenge CXOs have. Well, these are what people call microservices or serverless computing environments where they're breaking down the basic construct of an application to be a set of consumable services that can be orchestrated around particular data flows. And I think this is- You run into a problem of data. How do you discover those microservices? So having a trust that provided to go and aggregate all those microservices is a helpful approach. Guys, I know we're tight on time. You got to go and super thankful for your time coming on theCUBE and sharing your insight and color commentary on what's going on. Final question for both of you guys before you split is this. I've been watching NetApp for years, big fan of the company, obviously Silicon Valley Darling. Sometimes it takes a lot of heat. NetApp's dead and they never die, but you guys are always winning. Reinvention has been a big part of your culture. But that's not about pivoting. It's about building and just adjusting. Secret to the success. How do you guys do it? Advice for others? We have repeatedly leaned in to the thing that was going to kill us. So when VMware came along, everyone was like, oh, software defined data center. It was nobody's going to need data storage services anymore. Data management, VMware will do it all. And we said, you know what, that's not right. It's hard to do the data part and we're going to go make VMware better. And if we do that, our customers will pay us money to help them move to VMware faster. We leaned in on the thing that was going to kill us. And we're doing exactly the same thing. I mean, everyone was going, oh, Cloud's going to kill that. You built around it rather than let it roll over you. Not just built around it. We said, we'll make it better. And we did the same thing again with the Cloud. Oh, the Cloud's going to kill you. And we're like, you know what? Let's go figure out how to make Amazon better, make Microsoft better. If we can make them better, then I mean, if you solve a hard problem for a customer, somewhere or another, you can figure out how to get paid for that. And I think that's what we've been doing. And you get in early too. The timing is critical. It's not like you're late to the game and saying there's a pony in there somewhere. You look at it, although a little bit maybe, we first announced that we were working on this Cloud stuff three years ago. 2014, we had been started working in 2013. We were there from the ground with Amazon and with Azure running our on tap code. And they were changing their environment to fit with us. And we were changing our code to fit with them. And years later, when Microsoft says, who are we going to go to to help us manage the enterprise, they came to NetApp because we've been working with them for so long. I wish we had more time. We're going to get you in our studio in Palo Alto. Great conversation, real fire energy going on here from the execs here at NetApp. This is theCUBE, more live coverage in Las Vegas at NetApp Insight 2017 after this short break.