 from Berlin, Germany. It's theCUBE, covering DataWorks Summit Europe 2018. Brought to you by Hortonworks. Well, hello and welcome to theCUBE. We're here at DataWorks Summit 2018 in Berlin, Germany. And it's been a great show. And what we have now is we have IBM, specifically we have Dave Mectinel of IBM. And we're going to be talking with him for the next 10 minutes or so about Dave, you explain, you are in storage for IBM. And IBM, of course, is a partner of Hortonworks, who are of course the host of this show. So Dave, if you can introduce, give us your capacity or role at IBM. Discuss the partnership with Hortonworks. And really what's your perspective on the market for storage systems for big data right now and going forward? And what kind of workloads and what kind of requirements are customers coming to you with for storage systems now? Okay, sure. So I lead alliances for the storage business unit. In Hortonworks, we actually partner with Hortonworks, not just in our storage business unit, but also with our analytics counterparts, our power counterparts, and we're in discussions with many others, right? Our partner organization services and so forth. So the nature of our relationship is quite broad compared to many of our others. We're working with them in the analytics space. So these are a lot of these big data, data lakes, BDNA, a lot of people will use as an acronym. These are the types of workloads that customers are using us both for. And it's not new anymore. By now, they're well past their first half dozen applications. We've got customers running hundreds of applications. These are production applications now. So it's all about how can I be more efficient? How can I grow this? How can I get the best performance and scalability and ease of management to deploy these in a way that's manageable? Because if I have 400 production applications, that's not often any corner anymore. So that's how I would describe it in a nutshell. One of the trends that we're seeing at Wikibon, of course I'm the lead analyst for big data analytics at Wikibon under SiliconANGLE media. We're seeing a trend in the marketplace towards, I wouldn't call them appliances, but what I would call them is workload optimized, hardware, software, platforms, so they can combine storage with compute and are optimized for AI and machine learning and so forth. Is that something that you're hearing from customers that they require those built out AI optimized storage systems, or is that far in the future, or give me a sense for whether IBM is doing anything in that area and whether that's on your horizon? If you were to define all of IBM in five words or less, you would say artificial intelligence and cloud computing. So this is something that gets a lot of thought in mind share. So absolutely we hear about it a lot. It's a very broad market with a lot of diverse requirements. So we hear people asking for the converged infrastructure for appliance solutions. There's of course hyperconverged. We actually have either directly or with partners, answers to all of those. Now we do think one of the things that customers want to do is they're going to scale and grow in these environments is to take a software defined strategy so they're not limited. They're not limited by hardware blocks. They don't want to have to buy processing power and spend all that money on it when really all they need is more data. So there's pros and cons to the difference. You power AI systems, I know that. So that's under that broad heading, yeah. Yes, yes, yes. So of course we have packages that we've modeled in AI. They feed off of some of the Hortonworks data lakes that we're building. Of course we see a lot of people putting these on new pieces of infrastructure because they don't want to put this on their production application. So they're extracting data from maybe a Hortonworks data lake number one, Hortonworks data lake number two, some of the EDWs, some external data and putting that into the AI infrastructure. As customers move their cloud infrastructures towards more edge facing environments or edge applications, how are storage requirements changing or evolving in terms of in the move to edge computing? Can you give us a sense for the sort of trends you're seeing in that area? Well if we're going to the world of AI and cognitive applications, all that data that I might've thrown in the cloud five years ago, I now am educated enough because I've been paying bills for a few years on just how expensive it is and if I'm going to be bringing that data back, some of which I don't even know I'm going to be bringing back, it gets extremely expensive. So we see a pendulum shift coming back where now a lot of data is going to be on-house, sorry, on-premise, but it's not going to stay there. They need the flexibility to move it here, there or everywhere. So if it's going to come back, how can we bring customers, some of that flexibility that they liked about the cloud, the speed, the ease of deployment, even a consumption-based model. So these are very big changes on a traditional storage manufacturer like ourselves, right? So that's requiring a lot of development in software, it's requiring a lot of development in our business model and one of the biggest things you hear us talk about this year is IBM Cloud Private, which does exactly that. And it gives them something they can work with that's flexible, it's agile and allows you to take containerized based applications and move them back and forth as you please. Yeah, so containerized application, so if you can define it for our audience, what is a containerized application? You're talking about Docker and orchestrated through Kubernetes and so forth. So you mentioned Cloud Private, can you bring us up to speed on what exactly Cloud Private is and in terms of the storage requirements or storage architecture within that portfolio? Yes, absolutely. So this is a set of infrastructure that's optimized for on-premise deployment that gives you multi-cloud access, not just IBM Cloud, Amazon Web Services, Microsoft Azure, et cetera. And then it also gives you multiple architectural choices, basically wrapped by software to allow you to move those containers around and put them where you want them at the right time, at the right place, given the business requirement at that hour. Now as the data storage are persisted in the container itself, I know that's fairly difficult to do in a Dockering environment. How do you handle persistence of data for containerized applications within your architecture? Okay, some of those are going to be application specific. It's the question of designing the right data management layer depending on the application. So we have software intelligence, some of it from open source, some of which we add on top of open source to bring some of the enterprise resilience and performance needed. And of course, you have to be very careful if the biggest trend in the world is unstructured data. Well, okay, fine. It's a lot of sensor data. That's still fairly easy to move around. But once we get into things like medical images, lots of video, you know, HD video for K video, those are things which you have to give a lot of thought to how to do that. And that's why we have lots of new partners that we work with that help us with Edge of Cloud, which gives that on-premise like performance in really a cloud like setup. Here's a question on the left field. You may not have the answer, but I'd like to hear your thoughts on this. How is as blockchain and IBM's been making significant investments in blockchain technology, database technology, how is blockchain changing the face of the storage industry in terms of customer's requirements for storage systems to manage data in distributed blockchains? Is that something you're hearing coming from customers as a requirement? I'm just trying to get a sense for whether that's, you know, is it moving customers towards more flash towards more distributed, edge-oriented or edge-deployed storage systems? Okay, so yes, yes, and yes. Okay. So all of a sudden, if you're doing things like a blockchain application, things become even more important than they are today. Yeah. Okay, so you can't lose a transaction. You can't have storage going down. So there's a lot more care and thought into the resiliency of the infrastructure. If I'm buying a diamond from you, I can't accept the excuse that my $100,000 diamond, maybe that's a little optimistic, my $10,000 diamond or yours, the transaction's corrupted because the data's not proper. Or if I want my privacy, I need to be assured that there's good data governance around that transaction and that that will be protected for a good 10, 20 and 30 years. So it's elevating the importance of all the infrastructure to a whole different level. So, switching our focus slightly, so we're here at DataWorks Summit in Berlin. So where are the largest growth markets right now for cloud storage systems? Is it APAC? Is it the North America? Or where are the growth markets in terms of regions, in terms of vertical industries right now in the marketplace for enterprise-grade storage systems for big data in the cloud? That's a great question, because we certainly have these conversations globally. I'd say the place where we're seeing the most activity would be the Americas. We see it in China. We see it, we have a lot of interesting engagements and people reaching out to us. I would say by market, you can also point to financial services in more than those two regions. Financial services, healthcare, retail, these are probably the top verticals. I think it's probably safe to assume, and we can say that federal governments also have a lot of stringent requirements and requirements, new applications around the space as well. Right, GDPR, how is that impacting your customers' storage requirements or the requirement for GDPR compliance? Is that moving the needle in terms of their requirement for consolidated storage of the data that they need to maintain? I mean, obviously there's a security, but there's just the sheer amount of, is it leading to consolidation or centralization of storage of customer data that would seem to make it easier to control and monitor usage of the data? Or is it making a difference at all? It's making a big difference. Not many people encrypt data today, so there's a whole new level of interest in encryption at many different levels, data at rest, data in motion. There's new levels of focus and attention on performance, on the ability for customers to get their arms around disparate islands of data, because now GDPR is not only a legal requirement that requires you to be able to have it, but you've also got timelines that are within, which you're expected to act on a request from a customer to have your data removed, and most of those will have a baseline of 30 days, so you can't fool around now. It's not just a nice to have. It's an actual core part of a business requirement that if you don't have a good strategy for it, you could be spending tens of millions of dollars in liability if you're not ready for it. Well, Dave, thank you very much for the end of our time. This has been Dave McDonnell of IBM talking about system storage, and of course a big Hortonworks partner, and we are here on day two of the DataWorks Summit, and I'm James Kabilis of Wikibon SiliconANGLE Media, and have a good day.