>> Announcer: Live from San Jose, in the heart of Silicon Valley, it's theCUBE covering DataWorks Summit 2017. Brought to you by Hortonworks. >> Good morning, welcome to theCUBE. We are live at day 2 of the DataWorks Summit, and have had a great day so far, yesterday and today, I'm Lisa Martin with my co-host George Gilbert. George and I are very excited to be joined by a multiple CUBE alumni, the co-founder and VP of Engineering at Hortonworks Arun Murthy. Hey, Arun. >> Thanks for having me, it's good to be back. >> Great to have you back, so yesterday, great energy at the event. You could see and hear behind us, great energy this morning. One of the things that was really interesting yesterday, besides the IBM announcement, and we'll dig into that, was that we had your CEO on, as well as Rob Thomas from IBM, and Rob said, you know, one of the interesting things over the last five years was that there have been only 10 companies that have beat the S&P 500, have outperformed, in each of the last five years, and those companies have made big bets on data science and machine learning. And as we heard yesterday, these four meta-trains IoT, cloud streaming, analytics, and now the fourth big leg, data science. Talk to us about what Hortonworks is doing, you've been here from the beginning, as a co-founder I've mentioned, you've been with Hadoop since it was a little baby. How is Hortonworks evolving to become one of those big users making big bets on helping your customers, and yourselves, leverage machine loading to really drive the business forward? >> Absolutely, a great question. So, you know, if you look at some of the history of Hadoop, it started off with this notion of a data lake, and then, I'm talking about the enterprise side of Hadoop, right? I've been working for Hadoop for about 12 years now, you know, the last six of it has been as a vendor selling Hadoop to enterprises. They started off with this notion of data lake, and as people have adopted that vision of a data lake, you know, you bring all the data in, and now you're starting to get governance and security, and all of that. Obviously the, one of the best ways to get value over the data is the notion of, you know, can you, sort of, predict what is going to happen in your world of it, with your customers, and, you know, whatever it is with the data that you already have. So that notion of, you know, Rob, our CEO, talks about how we're trying to move from a post-transactional world to a pre-transactional world, and doing the analytics and data sciences will be, obviously, with me. We could talk about, and there's so many applications of it, something as similar as, you know, we did a demo last year of, you know, of how we're working with a freight company, and we're starting to show them, you know, predict which drivers and which routes are going to have issues, as they're trying to move, alright? Four years ago we did the same demo, and we would say, okay this driver has, you know, we would show that this driver had an issue on this route, but now, within the world, we can actually predict and let you know to take preventive measures up front. Similarly internally, you know, you can take things from, you know, mission-learning, and log analytics, and so on, we have a internal problem, you know, where we have to test two different versions of HDP itself, and as you can imagine, it's a really, really hard problem. We have the support, 10 operating systems, seven databases, like, if you multiply that matrix, it's, you know, tens of thousands of options. So, if you do all that testing, we now use mission-learning internally, to look through the logs, and kind of predict where the failures were, and help our own, sort of, software engineers understand where the problems were, right? An extension of that has been, you know, the work we've done in Smartsense, which is a service we offer our enterprise customers. We collect logs from their Hadoop clusters, and then they can actually help them understand where they can either tune their applications, or even tune their hardware, right? They might have a, you know, we have this example I really like where at a really large enterprise Financial Services client, they had literally, you know, hundreds and, you know, and thousands of machines on HDP, and we, using Smartsense, we actually found that there were 25 machines which had bad NIC configuration, and we proved to them that by fixing those, we got a 30% to put back on their cluster. At that scale, it's a lot of money, it's a lot of cap, it's a lot of optics So, as a company, we try to ourselves, as much as we, kind of, try to help our customers adopt it, that make sense? >> Yeah, let's drill down on that even a little more, cause it's pretty easy to understand what's the standard telemetry you would want out of hardware, but as you, sort of, move up the stack the metrics, I guess, become more custom. So how do you learn, not just from one customer, but from many customers especially when you can't standardize what you're supposed to pull out of them? >> Yeah so, we're sort of really big believers in, sort of, doctoring your own stuff, right? So, we talk about the notion of data lake, we actually run a Smartsense data lake where we actually get data across, you know, the hundreds of of our customers, and we can actually do predictive mission-learning on that data in our own data lake. Right? And to your point about how we go up the stack, this is, kind of, where we feel like we have a natural advantage because we work on all the layers, whether it's the sequel engine, or the storage engine, or, you know, above and beyond the hardware. So, as we build these models, we understand that we need more, or different, telemetry right? And we put that back into the product so the next version of HDP will have that metrics that we wanted. And, now we've been doing this for a couple of years, which means we've done three, four, five turns of the crank, obviously something we always get better at, but I feel like, compared to where we were a couple of years ago when Smartsense first came out, it's actually matured quite a lot, from that perspective. >> So, there's a couple different paths you can add to this, which is customers might want, as part of their big data workloads, some non-Hortonworks, you know, services or software when it's on-prem, and then can you also extend this management to the Cloud if they want to hybrid setup where, in the not too distant future, the Cloud vendor will be also a provider for this type of management. >> So absolutely, in fact it's true today when, you know, we work with, you know, Microsoft's a great partner of ours. We work with them to enable Smartsense on HDI, which means we can actually get the same telemetry back, whether you're running the data on an on-prem HDP, or you're running this on HDI. Similarly, we shipped a version of our Cloud product, our Hortonworks Data Cloud, on Amazon and again Smartsense preplanned there, so whether you're on an Amazon, or a Microsoft, or on-prem, we get the same telemetry, we get the same data back. We can actually, if you're a customer using many of these products, we can actually give you that telemetry back. Similarly, if you guys probably know this we have, you were probably there in an analyst when they announced the Flex Support subscription, which means that now we can actually take the support subscription you have to get from Hortonworks, and you can actually use it on-prem or on the Cloud. >> So in terms of transforming, HDP for example, just want to make sure I'm understanding this, you're pulling in data from customers to help evolve the product, and that data can be on-prem, it can be in a Microsoft lesur, it can be an AWS? >> Exactly. The HDP can be running in any of these, we will actually pull all of them to our data lake, and they actually do the analytics for us and then present it back to the customers. So, in our support subscription, the way this works is we do the analytics in our lake, and it pushes it back, in fact to our support team tickets, and our sales force, and all the support mechanisms. And they get a set of recommendations saying Hey, we know this is the work loads you're running, we see these are the opportunities for you to do better, whether it's tuning a hardware, tuning an application, tuning the software, we sort of send the recommendations back, and the customer can go and say Oh, that makes sense, the accept that and we'll, you know, we'll update the recommendation for you automatically. Then you can have, or you can say Maybe I don't want to change my kernel pedometers, let's have a conversation. And if the customer, you know, is going through with that, then they can go and change it on their own. We do that, sort of, back and forth with the customer. >> One thing that just pops into my mind is, we talked a lot yesterday about data governance, are there particular, and also yesterday on stage were >> Arun: With IBM >> Yes exactly, when we think of, you know, really data-intensive industries, retail, financial services, insurance, healthcare, manufacturing, are there particular industries where you're really leveraging this, kind of, bi-directional, because there's no governance restrictions, or maybe I shouldn't say none, but. Give us a sense of which particular industries are really helping to fuel the evolution of Hortonworks data lake. >> So, I think healthcare is a great example. You know, when we started off, sort of this open-source project, or an atlas, you know, a couple of years ago, we got a lot of traction in the healthcare sort of insurance industry. You know, folks like Aetna were actually founding members of that, you know, sort of consortium of doing this, right? And, we're starting to see them get a lot of leverage, all of this. Similarly now as we go into, you know, Europe and expand there, things like GDPR, are really, really being pardoned, right? And, you guys know GDPR is a really big deal. Like, you pay, if you're not compliant by, I think it's like March of next year, you pay a portion of your revenue as fines. That's, you know, big money for everybody. So, I think that's what we're really excited about the portion with IBM, because we feel like the two of us can help a lot of customers, especially in countries where they're significantly, highly regulated, than the United States, to actually get leverage our, sort of, giant portfolio of products. And IBM's been a great company to atlas, they've adopted wholesale as you saw, you know, in the announcements yesterday. >> So, you're doing a Keynote tomorrow, so give us maybe the top three things, you're giving the Keynote on Data Lake 3.0, walk us through the evolution. Data Lakes 1.0, 2.0, 3.0, where you are now, and what folks can expect to hear and see in your Keynote. >> Absolutely. So as we've, kind of, continued to work with customers and we see the maturity model of customers, you know, initially people are staying up a data lake, and then they'd want, you know, sort of security, basic security what it covers, and so on. Now, they want governance, and as we're starting to go to that journey clearly, our customers are pushing us to help them get more value from the data. It's not just about putting the data lake, and obviously managing data with governance, it's also about Can you help us, you know, do mission-learning, Can you help us build other apps, and so on. So, as we look to there's a fundamental evolution that, you know, Hadoop legal system had to go through was with advance of technologies like, you know, a Docker, it's really important first to help the customers bring more than just workloads, which are sort of native to Hadoop. You know, Hadoop started off with MapReduce, obviously Spark's went great, and now we're starting to see technologies like Flink coming, but increasingly, you know, we want to do data science. To mass market data science is obviously, you know, people, like, want to use Spark, but the mass market is still Python, and R, and so on, right? >> Lisa: Non-native, okay. >> Non-native. Which are not really built, you know, these predate Hadoop by a long way, right. So now as we bring these applications in, having technology like Docker is really important, because now we can actually containerize these apps. It's not just about running Spark, you know, running Spark with R, or running Spark with Python, which you can do today. The problem is, in a true multi-tenant governed system, you want, not just R, but you want specifics of a libraries for R, right. And the libraries, you know, George wants might be completely different than what I want. And, you know, you can't do a multi-tenant system where you install both of them simultaneously. So Docker is a really elegant solution to problems like those. So now we can actually bring those technologies into a Docker container, so George's Docker containers will not, you know, conflict with mine. And you can actually go to the races, you know after the races, we're doing data signs. Which is really key for technologies like DSX, right? Because with DSX if you see, obviously DSX supports Spark with technologies like, you know, Zeppelin which is a front-end, but they also have Jupiter, which is going to work the mass market users for Python and R, right? So we want to make sure there's no friction whether it's, sort of, the guys using Spark, or the guys using R, and equally importantly DSX, you know, in the short map will also support things like, you know, the classic IBM portfolio, SBSS and so on. So bringing all of those things in together, making sure they run with data in the data lake, and also the computer in the data lake, is really big for us. >> Wow, so it sounds like your Keynote's going to be very educational for the folks that are attending tomorrow, so last question for you. One of the themes that occurred in the Keynote this morning was sharing a fun-fact about these speakers. What's a fun-fact about Arun Murthy? >> Great question. I guess, you know, people have been looking for folks with, you know, 10 years of experience on Hadoop. I'm here finally, right? There's not a lot of people but, you know, it's fun to be one of those people who've worked on this for about 10 years. Obviously, I look forward to working on this for another 10 or 15 more, but it's been an amazing journey. >> Excellent. Well, we thank you again for sharing time again with us on theCUBE. You've been watching theCUBE live on day 2 of the Dataworks Summit, hashtag DWS17, for my co-host George Gilbert. I am Lisa Martin, stick around we've got great content coming your way.

>> [Announcer] Live, from Boston, Massachusetts, it's the Cube, covering Spark Summit East 2017, brought to you by Data Breaks. Now, your host, Dave Alante and George Gilbert. >> Welcome back to snowy Boston everybody, this is The Cube, the leader in live tech coverage. Arun Murthy is here, he's the founder and vice president of engineering at Horton Works, father of YARN, can I call you that, godfather of YARN, is that fair, or? (laughs) Anyway. He's so, so modest. Welcome back to the Cube, it's great to see you. >> Pleasure to have you. >> Coming off the big keynote, (laughs) you ended the session this morning, so that was great. Glad you made it in to Boston, and uh, lot of talk about security and governance, you know we've been talking about that years, it feels like it's truly starting to come into the main stream Arun, so. >> Well I think it's just a reflection of what customers are doing with the tech now. Now, three, four years ago, a lot of it was pilots, a lot of it was, you know, people playing with the tech. But increasingly, it's about, you know, people actually applying stuff in production, having data, system of record, running workloads both on prem and on the cloud, cloud is sort of becoming more and more real at mainstream enterprises. So a lot of it means, as you take any of the examples today any interesting app will have some sort of real time data feed, it's probably coming out from a cell phone or sensor which means that data is actually not, in most cases not coming on prem, it's actually getting collected in a local cloud somewhere, it's just more cost effective, why would we put up 25 data centers if you don't have to, right? So then you got to connect that data, production data you have or customer data you have or data you might have purchased and then join them up, run some interesting analytics, do geobased real time threat detection, cyber security. A lot of it means that you need a common way to secure data, govern it, and that's where we see the action, I think it's a really good sign for the market and for the community that people are pushing on these dimensions of the broader, because, getting pushed in this dimension because it means that people are actually using it for real production work loads. >> Well in the early days of Hadoop you really didn't talk that much about cloud. >> Yeah. >> You know, and now, >> Absolutely. >> It's like, you know, duh, cloud. >> Yeah. >> It's everywhere, and of course the whole hybrid cloud thing comes into play, what are you seeing there, what are things you can do in a hybrid, you know, or on prem that you can't do in a public cloud and what's the dynamic look like? >> Well, it's definitely not an either or, right? So what we're seeing is increasingly interesting apps need data which are born in the cloud and they'll stay in the cloud, but they also need transactional data which stays on prem, you might have an EDW for example, right? >> Right. >> There's not a lot of, you know, people want to solve business problems and not just move data from one place to another, right? Or back from one place to another, so it's not interesting to move an EDW to the cloud, and similarly it's not interesting to bring your IOT data or sensor data back into on-prem, right? Just makes sense. So naturally what happens is, you know, at Hortonworks we talk of kinds of modern app or a modern data app, which means a modern data app has to spare, has to sort of, you know, it can pass both on-prem data and cloud data. >> Yeah, you talked about that in your keynote years ago. Furio said that the data is the new development kit. And now you're seeing the apps are just so dang rich, >> Exactly, exactly. >> And they have to span >> Absolutely. >> physical locations, >> Yeah. >> But then this whole thing of IOT comes up, we've been having a conversation on The Cube, last several Cubes of, okay, how much stays out, how much stays in, there's a lot of debates about that, there's reasons not to bring it in, but you talked today about some of the important stuff will come back. >> Yeah. >> So the way this is, this all is going to be, you know, there's a lot of data that should be born in the cloud and stay there, the IOT data, but then what will happen increasingly is, key summaries of the data will move back and forth, so key summaries of your EDW will move to the cloud, sometimes key summaries of your IOT data, you know, you want to do some sort of historical training in analytics, that will come back on-prem, so I think there's a bi-directional data movement, but it just won't be all the data, right? It'll be key interesting summaries of the data but not all of it. >> And a lot of times, people say well it doesn't matter where it lives, cloud should be an operating model, not a place where you put data or applications, and while that's true and we would agree with that, from a customer standpoint it matters in terms of performance and latency issues and cost and regulation, >> And security and governance. >> Yeah. >> Absolutely. >> You need to think those things through. >> Exactly, so I mean, so that's what we're focused on, to make sure that you have a common security and governance model regardless of where data is, so you can think of it as, infrastructure you own and infrastructure you lease. >> Right. >> Right? Now, the details matter of course, when you go to the cloud you lose S3 for example or ADLS from Microsoft, but you got to make sure that there's a common sort of security governance front and top of it, in front of it, as an example one of the things that, you know, in the open source community, Ranger's a really sort of key project right now from a security authorization and authentication standpoint. We've done a lot of work with our friends at Microsoft to make sure, you can actually now manage data in Wasabi which is their object store, data stream, natively with Ranger, so you can set a policy that says only Dave can access these files, you know, George can access these columns, that sort of stuff is natively done on the Microsoft platform thanks to the relationship we have with them. >> Right. >> So that's actually really interesting for the open source communities. So you've talked about sort of commodity storage at the bottom layer and even if they're different sort of interfaces and implementations, it's still commodity storage, and now what's really helpful to customers is that they have a common security model, >> Exactly. >> Authorization, authentication, >> Authentication, lineage prominence, >> Oh okay. >> You want to make sure all of these are common sources across. >> But you've mentioned off of the different data patterns, like the stuff that might be streaming in on the cloud, what, assuming you're not putting it into just a file system or an object store, and you want to sort of merge it with >> Yeah. >> Historical data, so what are some of the data stores other than the file system, in other words, newfangled databases to manage this sort of interaction? >> So I think what you're saying is, we certainly have the raw data, the raw data is going to line up in whatever cloud native storage, >> Yeah. >> It's going to be Amazon, Wasabi, ADLS, Google Storage. But then increasingly you want, so now the patterns change so you have raw data, you have some sort of an ETL process, what's interesting in the cloud is that even the process data or, if you take the unstructured raw data and structure it, that structured data also needs to live on the cloud platform, right? The reason that's important is because A, it's cheaper to use the native platform rather than set up your own database on top of it. The other one is you also want to take advantage of all the native sources that the cloud storage provides, so for example, linking your application. So automatically data in Wasabi, you know, if you can set up a policy and easily say this structured data stable that I have of which is a summary of all the IOT activity in the last 24 hours, you can, using the cloud provider's technologies you can actually make it show up easily in Europe, like you don't have to do any work, right? So increasingly what we Hortonworks focused a lot on is to make sure that we, all of the computer engines, whether it's Spark or Hive or, you know, or MapReduce, it doesn't really matter, they're all natively working on the cloud provider's storage platform. >> [George] Okay. >> Right, so, >> Okay. >> That's a really key consideration for us. >> And the follow up to that, you know, there's a bit of a misconception that Spark replaces Hadoop, but it actually can be a processing, a compute engine for, >> Yeah. >> That can compliment or replace some of the compute engines in Hadoop, help us frame, how you talk about it with your customers. >> For us it's really simple, like in the past, the only option you had on Hadoop to do any computation was MapReduce, that was, I started working in MapReduce 11 years ago, so as you can imagine, it's a pretty good run for any technology, right? Spark is definitely the interesting sort of engine for sort of the, anything from mission learning to ETL for data on top of Hadoop. But again, what we focus a lot on is to make sure that every time we bring in, so right now, when we started on HTP, the first on HTP had about nine open source projects literally just nine. Today, the last one we shipped was 2.5, HTP 2.5 had about 27 I think, like it's a huge sort of explosion, right? But the problem with that is not just that we have 27 projects, the problem is that you're going to make sure each of the 27 work with all the 26 others. >> It's a QA nightmare. >> Exactly. So that integration is really key, so same thing with Spark, we want to make sure you have security and YARN (mumbles), like you saw in the demo today, you can now run Spark SQL but also make sure you get low level (mumbles) masking, all of the enterprise capabilities that you need, and I was at a financial services three or four weeks ago in Chicago. Today, to do equivalent of what I showed today on demo, they need literally, they have a classic ADW, and they have to maintain anywhere between 1500 to 2500 views of the same database, that's a nightmare as you can imagine. Now the fact that you can do this on the raw data using whether it's Hive or Spark or Peg or MapReduce, it doesn't really matter, it's really key, and that's the thing we push to make sure things like YARN security work across all the stacks, all the open source techs. >> So that makes life better, a simplification use case if you will, >> Yeah. >> What are some of the other use cases that you're seeing things like Spark enable? >> Machine learning is a really big one. Increasingly, every product is going to have some, people call it, machine learning and AI and deep learning, there's a lot of techniques out there, but the key part is you want to build a predictive model, in the past (mumbles) everybody want to build a model and score what's happening in the real world against model, but equally important make sure the model gets updated as more data comes in on and actually as the model scores does get smaller over time. So that's something we see all over, so for example, even within our own product, it's not just us enabling this for the customer, for example at Hortonworks we have a product called SmartSense which allows you to optimize how people use Hadoop. Where the, what are the opportunities for you to explore deficiencies within your own Hadoop system, whether it's Spark or Hive, right? So we now put mesh learning into SmartSense. And show you that customers who are running queries like you are running, Mr. Customer X, other customers like you are tuning Hadoop this way, they're running this sort of config, they're using these sort of features in Hadoop. That allows us to actually make the product itself better all the way down the pipe. >> So you're improving the scoring algorithm or you're sort of replacing it with something better? >> What we're doing there is just helping them optimize their Hadoop deploys. >> Yep. >> Right? You know, configuration and tuning and kernel settings and network settings, we do that automatically with SmartSense. >> But the customer, you talked about scoring and trying to, >> Yeah. >> They're tuning that, improving that and increasing the probability of it's accuracy, or is it? >> It's both. >> Okay. >> So the thing is what they do is, you initially come with a hypothesis, you have some amount of data, right? I'm a big believer that over time, more data, you're better off spending more, getting more data into the system than to tune that algorithm financially, right? >> Interesting, okay. >> Right, so you know, for example, you know, talk to any of the big guys on Facebook because they'll do the same, what they'll say is it's much better to get, to spend your time getting 10x data to the system and improving the model rather than spending 10x the time and improving the model itself on day one. >> Yeah, but that's a key choice, because you got to >> Exactly. >> Spend money on doing either, >> One of them. >> And you're saying go for the data. >> Go for the data. >> At least now. >> Yeah, go for data, what happens is the good part of that is it's not just the model, it's the, what you got to really get through is the entire end to end flow. >> Yeah. >> All the way from data aggregation to ingestion to collection to scoring, all that aspect, you're better off sort of walking through the paces like building the entire end to end product rather than spending time in a silo trying to make a lot of change. >> We've talked to a lot of machine learning tool vendors, application vendors, and it seems like we got to the point with Big Data where we put it in a repository then we started doing better at curating it and understanding it then starting to do a little bit exploration with business intelligence, but with machine learning, we don't have something that does this end to end, you know, from acquiring the data, building the model to operationalizing it, where are we on that, who should we look to for that? >> It's definitely very early, I mean if you look at, even the EDW space, for example, what is EDW? EDW is ingestion, ETL, and then sort of fast query layer, Olap BI, on and on and on, right? So that's the full EDW flow, I don't think as a market, I mean, it's really early in this space, not only as an overall industry, we have that end to end sort of industrialized design concept, it's going to take time, but a lot of people are ahead, you know, the Google's a world ahead, over time a lot of people will catch up. >> We got to go, I wish we had more time, I had so many other questions for you but I know time is tight in our schedule, so thanks so much Arun, >> Appreciate it. For coming on, appreciate it, alright, keep right there everybody, we'll be back with our next guest, it's The Cube, we're live from Spark Summit East in Boston, right back. (upbeat music)