 Live from the Fairmont Hotel in San Jose, California, it's The Cube at Big Data SV 2015. Welcome back everybody, this is Jeff Kelly, I'm with Wikibon and we are live on The Cube at Big Data SV 2015 in San Jose. We're kicking off day three of coverage and we're starting with the bang. We've got a great guest, Sunny Madra, who's a VP of data products at Pivotal joining us. Sunny, welcome to The Cube. Yeah, thanks for having me. So obviously there's been a lot of news announcements this week that we want to get to, but why don't we kind of level set, tell us a little bit about Pivotal, kind of the state of the big data business there, but also put in a little bit larger context because obviously Pivotal's also got the cloud component, the application development component. Give us an update on the business. Yeah, so 2014 was a really great year for us. Really, business really took off in a lot of different, all different aspects, and so when we look at the data business, we exited the business at $100 million in software bookings and one of the big things that we did last year was we created a product offering called the Big Data Suite and we offer that in a subscription model and that was a big change because typically we had our products offered individually, we offered them with perpetual licenses and we switched to the subscription model and that was really based on sort of looking at how our customers were using our products and how our customers were going on a particular journey. How they were going from say MPP to Hadoop and how they were layering on in-memory technologies and our customers really wanted like a packaged offering around that and so of the $100 million, $40 million was of our subscription product which we only launched in Q2 of last year so we had a really strong business and kind of the other thing that we saw together looking at the question that you said about the Cloud Foundry business and our agile businesses, we really see customers following this following journey with us and so they work with our big data groups to capture information. They then engage our data science teams to gain some insight from the information that they're capturing and capturing either that in structured databases on structured databases or in-memory, right? Once they get that insight now, this is where Cloud Foundry and Pivotal Labs really come in, they take that insight and they can work to operationalize that in a data-driven application that they can run on Cloud Foundry and so that's when, and that can be built with the help of Pivotal Labs and so that's like the businesses really come in together from all the way from the data all the way to data-driven applications so that's been really exciting for us. Yeah, so you guys really try to cover the whole spectrum from capturing the data, analyzing it for insight and then operationalizing those insights in the form of applications. Yeah, yeah. So let's get to some of the news this week, a couple of big announcements. The first one I wanted to talk about was a decision on Pivotal's part two, open source your big data products, your suite of products. Yeah. That includes Green Plum, Hawk, which is your- SQL on Hadoop. SQL on Hadoop Engine, Gemfire and some other big data products. Talk about that, what was the decision, what was the internal discussions about that and what was the impetus for that decision? Yeah, so one, looking at the success that we had with Cloud Foundry, so there was a, we published some numbers on Cloud Foundry as well and we had real tremendous success last year and looking, having gone through that process because that was a project that started inside VMware that then became an open source, right? But looking at the success of commercializing that and building a community around it really gave us confidence that there is an ability to have open source software that has commercial distributions and the business model is not just support, right? And so by having, by seeing that unfold it really gave us a confidence to embrace this larger industry trend which is open source. And so the way I really think about that is one, we're starting to see open source show up in RFP requirements. Is there an open source component to this, right? Is to the software that I'm buying and that's because vendors don't want to be, customers don't want to be locked into vendors and so that's a big driving force. The second one, which again, we saw in Cloud Foundry is given the success of the large internet companies and then the new generation of companies, when enterprises look at those companies, how are these companies creating so much value quickly? And then when they kind of peel underneath the hood they're saying, wow, open source is there. But those companies are not just using open source software, they're creating open source software and they're contributing to open source software. So what we saw in Cloud Foundry is customers are buying the software from us but they're also showing up and working with us and saying, hey, look, there's just something on the roadmap for Cloud Foundry. You have it nine or 12 months out. I need it now, right? I need to solve this falling business problem so that I don't get leapfrog by one of my competitors or a new upstart. And the way for them to do that is to get involved in the open source. And so that's a really big change that we're starting to see there where they're buying this software and saying, hey, we want to send teams to also contribute so that we can get this feature done and accelerated or built the way we need it. And so that was a really big driving force that we saw in the Cloud Foundry project. And we've always had interactions with customers where we open up our roadmaps to them and we let them have access to it but we never really allowed them to contribute. And then lastly, the way I like to describe it is the three of us were starting a company today and we wanted to solve a particular problem. We'd go to the open source, right? We'd go and look for particular projects. And when we look at those projects, we'd find the ones that had the most contributors and activity in the mailing list and that's what we'd use to start building our company. And so that's the new decision-making process, right? It's no longer the idea from 30 years ago where we'd go and we find the big companies, call them up, get their data sheets, look at it and say, it's open source. That's how we would go. And so we really want to embrace that trend and the trend of people working on open source and bringing that software in also internally, right? So we cut out the shadow sales for us, right? Where somebody's working on it and say, I'm working on this really cool project. We should use this at work. Those trends are just really big. Yeah, I think you hit on a few things that I want to bring my co-host into the conversation, Jeff Frick, that Jeff, you and I were talking about on the intros, the way open source is changing. You know, not just the enterprise business from a vendor perspective, but also the way practitioners are developing so much of the value themselves by they're innovating within the open source project themselves and that's really changing the business. Yeah, I'm just curious on the RFP. It's an interesting example. When they're looking for open source, how many of those times do they already have experience in an open source component of what you're going to put in and they want to continue on that path or how much of it is, you know, they want to start to develop the expertise and become part of the community because they know they could potentially innovate faster along a different path or, you know, bring things to bear without necessarily having to wait for your guys roadmap to catch up to the requirements. Yeah, so you know, I think it's been a journey, right? I think the first part of the journey was, you know, just embracing open source software. And we see that now like with Hadoop, right? Enterprise are embracing Hadoop. So that was the first part of the journey that really, you know, broke open. And now what we start seeing is that's why they, you know, the purchasing departments want that in there, right? They're saying, hey, look, we want the ability to have the flexibility not to be locked into any one vendor and have the ability for our teams to, you know, accelerate, right? If we, you know, and it's a good question for them to ask, if do you really want to be buying proprietary software? Right? Will this enable our business going forward? And so that's something that, you know, even like through the purchasing departments, procurement departments, they understand that they should ask that question, because this will limit their business in a lot of ways. Yeah, and it's really too almost a reflection of more of an agile, go-to-market methodology in the company, not software development, but just we don't know what we want all together now. Let's get going, get it in. And we want the ability to be able to move in real time and develop as well. Interesting that it's actually showing up in the RFPs. It is. And it's really, like I said earlier, it's all being driven by what's happening in the consumer internet space, right? You know, you see a company like Uber, right? Go to zero to $40 billion so quickly. And then you look back and say, these guys, they don't produce any physical good or anything like that. It's all being done in software, right? And it's all really embracing new modes of customer engagement. And so one of the things that we really think of, going forward, data is driving that customer engagement model. And so it's advertising and things like that. Those things are going to change. It's really going to be about that. Consumer internet companies really don't advertise. Uber doesn't advertise. They just gave a really good experience to their user and that really drives their business. And so when other more traditional enterprises are looking at that, they're really seeing that this is the area that they have to invest in. It's interesting too, your comment about the suite and subscription, because we're in San Jose, one of the biggest buildings in San Jose is Adobe. And they had very expensive products that you had to buy single licenses and they flipped to the creative suite subscription model. And I think it's been transformative. And I think you said a huge portion of your reported numbers are coming through this kind of new methodology to buy. I wonder if you can talk a little bit about why that's more receptive. How does this kind of map to their journey and is that really the way to do it? Adobe example is good, but it's exactly that. Customers are, they're transforming, right? We're in this age of digital transformation or a lot of things are changing. We're kind of moving away from the offshoring, outsourcing model, companies are realizing like soft force eating the world, right? And so they're really embracing agile development and bringing development on board. But as they do that, they need to have flexibility in the tools, right? So the challenge for them is if they buy a perpetual license for one product and their business requirements change over six to nine months, they don't want to have to show up and say, well, I have to go buy another license for something. And that's why the subscription model and then specifically the subscription model in a suite, which is where we are with Pivotal, big data suite is great for us, right? Where we end up giving the customers a flexibility to have structured data with our green plum product, right? Unstructured in our Hadoop distribution, right? You know, we offer in memory with Gemfire. What we've also added this year to the big data suite is we've also added Redis and RabbitMQ and a Cloud Foundry Foundation. So that really allows companies additional entitlements to not only do the data capture side of things, but also to move to operationalize that in applications. That's the one piece that we didn't have in the last year. So now we've added that into the suite so we can kind of. So you've really flipped the sales and delivery methodology based on the way that they want to buy in their journey as opposed to trying to make them fit into what you've got to go sell. Well, yeah, exactly, right? And I think just, you know, as we see customers wanting to operationalize data, we can create data-driven applications. We don't want to have them keep coming back to say, look, buy this suite from us. You can go from capturing data all the way to creating an application and you can move your licenses around as you need to. So what are the implications for Pivotal's big data business from a revenue perspective? So if you're open sourcing a lot of the database, where are you going to monetize? Is it higher up the stack? How are you going to tackle that? Yeah, good question. So, you know, like I said, again, leveraging Cloud Foundry for us is we're going to have commercial distributions. And those commercial distributions will retain some enterprise functionality that we only offer in the commercial distributions. And so when we think of the specific products like in Gemfire, right? So Gemfire will offer it open source. Product will offer it be fully functional. But when you want something like, say, WAN replication or continuous querying technology, you know, that really for disaster recoveries or large clusters or cluster multiple data centers, that's the B of the functionality available in our enterprise distribution. Similarly with our, you know, our hawk technology, you know, we have a next generation query optimizer. So we're going to put our current generation query optimizer, but our next one, we'll make that available in the commercial distribution. As we make the next next generation, we'll push that one into the open source. Very good. So of course, the other big news this week was the open data platform. So, I've got several questions about that, but why don't you start with just telling us what it is? Yeah, so, you know, really it's an industry collaboration. It's an initiative led by a number of folks. You know, we announced on Tuesday as about 15 people that have come in and we're getting more and more requests every day. So we're really excited about that. Really, you know, what we saw with our customers and with potential customers that we had a challenge where, sorry, they were challenged where they were using, you know, multiple distributions for multiple companies in order to solve their business problems. That was because, you know, there was a lot of fragmentation in the space. And the way I really like to think about it is think of Unix before Linux, right? And so, you know, where you had a bunch of different companies that had their own distribution and the ecosystem of software was sort of, you know, moved slower because there was no standardization. Linux comes around, you get a common kernel and then you get a situation where you have something like REL, CentOS and Oracle Linux. And if your software runs on any one of them, it'll run on the other two, right? And so that really enabled the ecosystem of software on Linux to grow and Linux to really take off. And so, you know, looking at that, we felt as though there needed to be some standardization around a common core where, you know, a bunch of companies can agree upon creating a distribution from there. So it's not really about contribution, you know, as like, you know, some of the posts have come out this week. It's really just about saying, hey, let's agree to a common core and all build their distribution from there. ASF should continue to manage, you know, the contribution, the governance of those projects and the direction. We're just saying, let us pick these certain versions and all agree upon it together and then create our distributions from there. And I think that makes it better for the, you know, the customers buying the software. It makes it better for the ecosystem of developers creating software on top of a dupe, you know, which I think this is one of the reasons we haven't seen that take off as much as it could. It's a challenge today, right? If you have to certify against three or four different vendors. Well, right, because as we've covered on theCUBE, Jeff, where all the big data applications is one of the questions we ask a lot. And there's not too many companies out there that are focused on building those kind of applications. Is this an effort to kind of address that issue? Creating a common core so people can feel comfortable if I build an application on this distribution, I can move it to this, that distribution. And do you think that will open up the big data application market? You hit it on the head, right? You see big data applications created today within, you know, companies, specifically within internet companies, because they don't have to fight the distribution war, right? Like if you're a, you know, if you're Facebook or something, you know, you have your own distro, right? And so then you can create a set of applications. You don't have to worry about it. But if you're an enterprise, you know, you're buying different products from different folks or you have existing products, right? And you need it to integrate with this, you know, new data layer that you have. And so exactly, you know, it's really what we want to accelerate. We want to accelerate the ecosystem of software that's being built on top of a dupe, which is, you know, it's there, but it's not as big as it, you know, we feel it could be. Well, you mentioned fragmentation. I wonder if you could give a couple of examples of what you mean by that. I was reading a post by Roman Shapochnik, who's your director of open source epitode. He was mentioning something that he referred to as kind of many ecosystems emerging around kind of the different distributions. And I think that's what you're referring to around fragmentation. Can you kind of add a little color to that? What's some of the things you're seeing happening that you're hoping to overcome with the ODP? Yeah, I mean, you know, first and foremost, like, you know, when we work with customers, right? We see customers, you know, running, you know, two to three different versions of HDFS within their enterprise, right? And that's, you know, one, and how that, let's really, you have to think about how that happens, right? So one group goes and decides to purchase, you know, Distro A from a particular vendor, right? And that's really made it easy for the other group to, you know, probably purchase Distro A, the, you know, MSA is already in place. They can just go and say, let's just expand that contract. But when Group B realizes, oh, we need this particular software, and that particular software is not certified on Distro A. So, oh, I got to go get Distro B, right? And then, you know, Group C shows up and says, hey, you know what, I need this other thing, and it's not certified on Distro A or B, I need this other thing. And so, you know, you're, we're seeing, you know, and a lot of customers or customers are running multiple versions, you know, not by design, but just by the fact of how the ecosystem is playing out right now. And so, you know, we feel as if there's a common core, you know, you wouldn't have that problem with that, you know, with these different groups. So, well, talk a little bit more about the relationship that ODP will have with the Apache Software Foundation. And how, what will that relationship be like? Because some of the criticism coming is that while we've got this group, ASF, that is tasked with doing some of the things you mentioned, how will they interact? How do you view the relationship? Yeah, so, you know, one of the key things with the, it's very complimentary, right? But one of the key things that the ASF has done extremely well, and I think they need to continue to do that. ASF doesn't generally recognize vendors, right? Or corporations. It's really around individual contributors. And I think that's the right thing, right? That's how the model works. That's how, you know, you contribute there. That's how the, you know, how you build, you know, your credibility within the ecosystem, right? Not just by being a big company, right? I think where that starts to have some challenges and where the ODP comes in and is complimentary, say, here's a bunch of companies that need to collaborate, not individual developers on a particular project, right? To say, you know, so the developers can continue to collaborate on what the next generation of features are of, you know, for HDFS, right? But when three different vendors have to get together, you know, we have to say, well, look, we should all use this version, x.y of HDFS in our distribution, right? And if we all do that, then the software that utilizes HDFS will, you know, be much easier to certify. So I think that's that complimentary aspect that, you know, we really bring to the ASF. The other interesting thing that I found about the Open Data Platform is it's not just vendors selling software. It's practitioners as well. You've got GE involved. I think Verizon is one of their initial members as well. What role will practitioner organizations, such as those play in the ODP? And that's by design, right? If it was just, you know, vendors, we felt as though it wouldn't be, so the idea is that, you know, the group of folks that make up the members are the ones that vote on the versions of the projects that need to be created. And it also allows on what additional projects we want to bring to the core, right? We started with a small core today, HDFS, Yarn, MapReduce, and Ambari. But as we want to expand that, you know, that group will work together. And that's why it's important to have the practitioners in there, because they have their own views based on, you know, the problems that they're trying to solve. Right. So we're running close on time, so I want to give you kind of the last word. So what's on the roadmap to the extent that you can share for both Pivotal, the ODP? I mean, what are you looking forward to in the next six, 12 months in terms of what excites you? What do you think we'll be talking about maybe in the next year at this table? Yeah, I think really, you know, we're, you know, I have a lot of the teams focused as, you know, the integration of cloud and our data products, really. And so that's been an area where I still think it's very early, and we're going to push there. And so you're going to see things coming out of us where, you know, much more, you know, I talked about a journey of, you know, a customer going from data capture to operationalizing it. We really think that the data products enabled in Cloud Foundry will be really interesting for us. So you start seeing a lot of that. And that really starts opening things up as you have the ability to push the products out of, you know, sort of the EDW groups and into the, you know, line of business groups to, you know, kind of really use those products as Cloud Foundry makes them available. Oh, I agree. I think that's where it's going to get interesting is when you start to move these, start to operationalize big data. And you've got more business people making decisions on these kind of analytics. That's where a lot of the value is coming. And that's where the ecosystem of applications will really start to, you know, where you're saying they're missing today. That's when they start coming. Right, I think we're all looking forward to that. So any monitor VP of data products at Pivotal. Thanks so much for joining us on theCUBE. All right. We appreciate it. Thanks everybody for watching. Stick around. We'll be right back with our next segment, I'm San Jose.