 Hi everybody, this is Jeff Kelly from wikibon.org. We're back here live at the HP Vertica Big Data Conference in Boston. My colleagues Dave Vellante and John Furrier are off doing some panel discussions right now. I'm your host for this segment and we're going to talk about some really interesting topic, the kind of the relationship between HP, which of course is putting a lot of effort and money into their big data strategy and a key partner of theirs, Cloudera, and how Hadoop is playing an important role in their big data platform known as Haven, of course Hadoop is the H in Haven. So we're welcoming Tom Picney, Director of Global Partner Sales at Cloudera. Welcome to the Cube for the first time and I appreciate you coming on. Thanks for having me. Absolutely. And from HP we've got Steve Tremac, Director of Big Data Systems Engineering, Converge Appliance Systems and Solutions, Converge Systems. That's a mouthful, but it's an important role. Thanks, Jeff. Great to be here. Yes, Steve. Thank you so much for coming on. So guys, yeah, why don't we start, Steve, with you. Tell us a little bit about your role at HP and then we kind of go into the relationship a little bit. Sure. So I'm part of, as you mentioned, a new business unit in HP called Converge Systems. This is a business unit focused on building appliances and reference architectures specifically around our Converge infrastructure, focused and purpose-built around workloads. So clearly a lot of excitement, a lot of energy here around Haven as the big data core platform for HP and my specific team is responsible for building appliances around the Haven stack. So we today have solutions, a building block solution for Vertica and we also have an appliance that we deliver to market jointly with our partner, Cloud Error. Yeah, we spoke with Paul Miller in the past about the strategy around and the thoughts around this, but why don't you kind of, in your own words, what is really the value proposition when it comes to the app system for Hadoop and kind of packaging that in that appliance format versus some of the other ways you can deploy Hadoop? Yeah, so when someone deploys a complex set of infrastructure, servers, stores, networking, software stack for multiple vendors, there are a few different ways they can approach it. Typically, an organization will take that on themselves, so they'll work with a partner and try to build something custom for their environment. And for that, we have products, we've got our Converge infrastructure products that allow them to do that. And that is, the success rate of that is directly dependent upon the skill set at the organization or the consultant and how well they've identified their needs. Well, in doing that, the engineers in our organization are focused on solutions. We have developed best practices. We understand how workloads like Hadoop, like Vertica, respond and the needs, specifically for balancing hardware and software and tuning and configuration. So from that, we built reference architectures. The thing about like a map that you might get from Avis, right? You get this map and it tells you how to get to the city, but to get to the street address is kind of tough, right? It gets you in the right neighborhood and a reference architecture documents some of those best practices. But what we're doing now with our appliances called App Systems, it's more like a GPS. It gets you right to the door and it takes the complexity out of building the infrastructure underneath. So what we do is we take some specific optimized servers and storage and networking. We pre-install Red Hat and the Cloudera distribution, you know what to do. We work with the customers to gather their intent so we configure it in the factory and it ships, racked and stacked and pre-installed, delivered to the customer site either through one of our partners or through HP. And the experience, the goal is that on day one, the environment is live in the customer network. So Tom, tell us a little bit about from Cloudera's perspective. What does the HP relationship and the role, you know, HP's expertise in the kind of packaging the infrastructure, how does that align with what Cloudera is trying to do and your customer base? Sure. Well, obviously HP market leadership in the X86 server market. So they've got a lot of expertise there and know that platform very, very well. So we're obviously not experts in hardware. And so it's great collaboration with HP to understand those platforms in great, great detail, both on the DL380 as well as the SL4540, which they came out with the reference architecture earlier this year. So, you know, it's great to be able to walk into an account and have both that infrastructure expertise as well as the Cloudera Hadoop expertise with those customers. And then obviously going in with HP, it's a tremendous partner, fantastic relationships within accounts and, you know, tremendous trust and kind of long standing partnerships with major customers for now adopting Cloudera. And well, and certainly in the Hadoop environment, well, I'd like to get both your opinions on this, but, you know, in the Hadoop environment, certainly the partnership success of anybody in that market is critical. It's a vibrant ecosystem. There's a lot of different players. Maybe just taking a step back. How does Cloudera kind of look at their partnering strategy? And again, I mean, and how that aligns with what HP is doing and why that's such a good fit? Sure. Well, you know, I think that customers want solutions. So they're not buying pieces parts. And when you look at, people are not trying to deploy Hadoop. They're not trying to deploy big data. They're trying to solve business problems. And when you try and solve those problems, not only do you have Cloudera as the Hadoop distribution, not only do you have HP as the hardware and infrastructure provider. There are a lot of other things that you need to provide. And so, you know, as Cloudera, our ability to bring forward certified partners, we've got over 700 partners. So our ability to have that ecosystem, to really provide solutions to solve those business problems. I think that's really powerful, really powerful for customers, because that's really what they're looking for. Right. And Steve, from your perspective, I mean, we heard in both keynotes yesterday from Colin Mahoney, VP and GM of Vertica. And of course, George Kedifa, who's running software for HP, mentioned that, you know, despite HP's size, the biggest technology company in the world for them to really, for HP to really leverage and take advantage of the big data opportunity, they need to partner with outside, outside companies. So talk a little bit about, from your perspective, how that, how you kind of go about looking at partnerships and, you know, specifically Cloudera, but even in more generally, kind of, what's your strategy? Yeah. And I think it comes back to Haven, right, that the Haven strategy is built on open source Hadoop is built on on standards. That's a critical part. We have an open partnering strategy and have reference architectures today with a number of distributions around Hadoop. And the end of Haven really is, is that's a number of solutions, right? So it's a critical as we're providing a foundation for, for ingesting and analyzing and, you know, and garnering the value out of out of multiple sources of big data. It's also critical that to complete the solutions for different industry segments or different customers that we work with, with best debris partners. And we provide a platform and a foundation to enable that for our customers. So, you know, we have, we have a strategy of open partnership and specifically with Cloudera, we have, we've gone even a step further with the reseller agreement. So we're able to provide the Cloudera bits and licensing as part of the solution we deliver to the customers. So not only are we bringing together best debris partnerships as part of the Haven ecosystem, but we're also able to deliver that easily to customers to take some of the complexity and the risk out, the fact that we're able to work together to implement the best practices around configuring and deploying this and roll that in codified, if you will, into the app system helps our customers be more successful when they're jumping into the space. So really lowering the barriers to entry and also taking away kind of de-risking the investment in a lot of ways. So I want to talk a little bit about customers and any trends you might be seeing in terms of adoption and use cases, because of course, Hadoop is still a relatively young technology, it's evolving quickly. You know, we hear about different types of use cases in the enterprise. I'm curious, Tom, from your perspective, what are you seeing in terms of your, you know, when you're going into new accounts these days? What are some of the, some of the use cases and some of the, some of the challenges of business problems you're seeing customers that they want to solve using Cloudera and Hadoop? Sure. Yeah, you know, it's interesting because I think if we were here 12 months ago or 18 months ago, we'd be talking more about advanced analytics, we'd be talking more about kind of web properties, we'd be talking about more kind of early adopters. And now I think we're really seeing enterprises trying to understand how to adopt Cloudera and how to adopt Hadoop. And so I think, you know, to that end, the Haven announcement is very powerful because it's HP helping organizations to figure out how to adopt this technology. And to that end as well, I think, you know, the app system is very powerful because it, it minimizes some of the kind of moving parts that customers need to understand as they adopt this. You know, specifically as we go into accounts, you know, rather than that kind of advanced analytics that you would have seen 12 or 18 months ago, we're really seeing a couple of really kind of basic operational efficiency kind of use cases. We see that organizations are looking at data transformations they're doing, either on their ETL grid, or transformations they're doing in their data warehouse. And typically they've got a couple which are quite expensive. And Cloudera is an optimal platform to be able to offload those, to free up capacity in their ETL grid and within their data warehouse for transformations. You know, kind of the second thing that we see is organizations have got additional detail that they want to keep in their data warehouse. And so that could be, you know, detailed records in out years, which is very expensive. But there is a business analyst somewhere who wants to run reports against them at some point. So, you know, you can put, keep them in a sort of high priced, high cost per terabyte data warehouse platform, or you've got the option of writing them off to storage. And so Cloudera provides an excellent kind of balance between those, you know, dramatically lower cost for that class of storage. Yeah, absolutely. We're seeing, you know, we're seeing the same things about members of the Wikibon community talking to us about some of the use cases. And certainly it's a Hadoop is definitely disrupting the traditional enterprise data warehouse market where you've got, you know, often scale up infrastructure, you've got proprietary hardware, you've got a very expensive solutions and people are looking to not remove those, but to kind of use Hadoop to to really make better use of that investment and awful at some of that work and lower some of those costs. Yeah, by all means. Definitely a compliment. Yeah. And Steve, what are you seeing? I think in looking at some of those different use cases. And one thing that's pretty exciting are some of the advancements on the converged infrastructure in the hardware front, particularly, you know, a couple of examples are SL4500 series platform provides in a 4.3 U space. It gives you options for a single compute platform with 60 spindles. It gives you options for two compute platforms each with 25 or three with 15. So if you think about what we have now are very, we have different configuration options that might fit some of the different use case models around being a true archival or data lake type of approach where we're just density of storage is the critical element. We have different different options that I think marry up with the technology. The other technology is Moonshot and looking at as the data continues to scale and some of the in the keynote, they talked about just the moving forward with how quickly machine and human data is scaling in the future that the amount of space and power necessary to store that is astronomical. Well, something like Moonshot starts to change that dynamic around compute and power density and saving. So it's a pretty exciting time. We're seeing a lot of marriage technologies here that I think fuel this next the next phase of this group. So I'm curious. So from customers, what are you hearing in terms of you similar to what's almost saying some of those use cases? What's the trend you're seeing? First summer, I think that's where where Haven really comes into play is that that the the marriage of of starting to look at, you know, the blending and movement of data, for example, from from Hadoop and HGFS into into Vertica for the speed of analytics and some of the blending of technologies and being able to to mine data out with things like like autonomy. It's we're seeing a maturation of the technology and the market and broader use cases. So certainly broader use cases and what are some of the I guess one of the issues we're looking to explore at Wikibon is really teasing out some of the real key characteristics of enterprise level big data platform. I mean, for you, from your perspective, from what you're seeing, what are some of those key areas? Is it security? Is it some of the hardware efficiency that we've been talking about? What are some of the keys to you? Yeah, I think that the the key is certainly as we look at security, particularly at the the discussion this morning around ArcSight and some of the as we continue to to leverage more and more the state of the organization at the trend for vulnerabilities that continues to grow. So security is a key part of that and also finding a way to efficiently scale the scale as the data needs grow within an organization. Scaling, you know, we're hearing, you know, petabytes not being thrown away that gigabytes and megabytes are thrown around just a few years ago, right? So so being able to effectively scale and balance out compute and capacity with power needs and space needs in the data center are the things that and being able to manage those large environments are key things that HP is on the head. And so Tom, in terms of, you know, from Cloudera's perspective, we know Cloudera certainly has been is not, you know, slowing down at all in terms of development of some of the some of the products and the new solutions you're you're releasing. We've got Century focused around security, Cloudera search earlier this summer, really enabling from the Cloudera search perspective an apology, just new ways to basically interact with the data that's in HTFS. So, you know, as you look forward, how does that make your job easier as the as you're in charge of, you know, this relationship with HP, a critical relationship to Cloudera? How does all the work that your engineering team is doing kind of make your job maybe a little bit easier and kind of improve that relationship? Well, you know, I think it goes back to the theme of the HP relationship is especially valuable as we go into the enterprise and going through the traditional old school mainstream enterprise, which is really getting pushed now. Those guys are getting pushed to have a big data strategy and, you know, nine times out of ten, a big data strategy, you know, is going to include a Hadoop strategy is sort of the corner tenant there. And so I think is Cloudera continues to develop its product and address enterprise requirements, such as security, such as multiple access points and ways to access that data. It just becomes a more and more solid data management infrastructure for the enterprise. And so it's just easier to go and sell to those CIOs of mainline kind of old school industries when there is functionality, which is kind of analogous to what they're used to having. All right, great. Well, we look forward to continued development between Cloudera and HP. We look forward to what you guys are going to bring to the table. Already doing some really exciting things. Steve, Tom, thank you so much for coming on theCUBE. Appreciate it. Again, first time on theCUBE. We hope you come back. You're now a CUBE alumni as we like to call it. So everybody, please stay with us. We've got a lot more coverage coming from here in Boston at the HP Big Data Vertical Conference. Stay right there and be right back.