>> Live from San Jose, in the heart of Silicon Valley, it's theCUBE, covering DataWorks Summit 2018. Brought to you by HortonWorks. >> We are wrapping up Day One of coverage of Dataworks here in San Jose, California on theCUBE. I'm your host, Rebecca Knight, along with my co-host, James Kobielus. We have two guests for this last segment of the day. We have Sudhir Hasbe, who is the director of product management at Google and Ram Venkatesh, who is VP of Engineering at Hortonworks. Ram, Sudhir, thanks so much for coming on the show. >> Thank you very much. >> Thank you. >> So, I want to start out by asking you about a joint announcement that was made earlier this morning about using some Hortonworks technology deployed onto Google Cloud. Tell our viewers more. >> Sure, so basically what we announced was support for the Hortonworks DataPlatform and Hortonworks DataFlow, HDP and HDF, running on top of the Google Cloud Platform. So this includes deep integration with Google's cloud storage connector layer as well as it's a certified distribution of HDP to run on the Google Cloud Platform. >> I think the key thing is a lot of our customers have been telling us they like the familiar environment of Hortonworks distribution that they've been using on-premises and as they look at moving to cloud, like in GCP, Google Cloud, they want the similar, familiar environment. So, they want the choice to deploy on-premises or Google Cloud, but they want the familiarity of what they've already been using with Hortonworks products. So this announcement actually helps customers pick and choose like whether they want to run Hortonworks distribution on-premises, they want to do it in cloud, or they wat to build this hybrid solution where the data can reside on-premises, can move to cloud and build these common, hybrid architecture. So, that's what this does. >> So, HDP customers can store data in the Google Cloud. They can execute ephemeral workloads, analytic workloads, machine learning in the Google Cloud. And there's some tie-in between Hortonworks's real-time or low latency or streaming capabilities from HDF in the Google Cloud. So, could you describe, at a full sort of detail level, the degrees of technical integration between your two offerings here. >> You want to take that? >> Sure, I'll handle that. So, essentially, deep in the heart of HDP, there's the HDFS layer that includes Hadoop compatible file system which is a plug-able file system layer. So, what Google has done is they have provided an implementation of this API for the Google Cloud Storage Connector. So this is the GCS Connector. We've taken the connector and we've actually continued to refine it to work with our workloads and now Hortonworks has actually bundling, packaging, and making this connector be available as part of HDP. >> So bilateral data movement between them? Bilateral workload movement? >> No, think of this as being very efficient when our workloads are running on top of GCP. When they need to get at data, they can get at data that is in the Google Cloud Storage buckets in a very, very efficient manner. So, since we have fairly deep expertise on workloads like Apache Hive and Apache Spark, we've actually done work in these workloads to make sure that they can run efficiently, not just on HDFS, but also in the cloud storage connector. This is a critical part of making sure that the architecture is actually optimized for the cloud. So, at our skill and our customers are moving their workloads from on-premise to the cloud, it's not just functional parity, but they also need sort of the operational and the cost efficiency that they're looking for as they move to the cloud. So, to do that, we need to enable these fundamental disaggregated storage pattern. See, on-prem, the big win with Hadoop was we could bring the processing to where the data was. In the cloud, we need to make sure that we work well when storage and compute are disaggregated and they're scaled elastically, independent of each other. So this is a fairly fundamental architectural change. We want to make sure that we enable this in a first-class manner. >> I think that's a key point, right. I think what cloud allows you to do is scale the storage and compute independently. And so, with storing data in Google Cloud Storage, you can like scale that horizontally and then just leverage that as your storage layer. And the compute can independently scale by itself. And what this is allowing customers of HDP and HDF is store the data on GCP, on the cloud storage, and then just use the scale, the compute side of it with HDP and HDF. >> So, if you'll indulge me to a name, another Hortonworks partner for just a hypothetical. Let's say one of your customers is using IBM Data Science Experience to do TensorFlow modeling and training, can they then inside of HDP on GCP, can they use the compute infrastructure inside of GCP to do the actual modeling which is more compute intensive and then the separate decoupled storage infrastructure to do the training which is more storage intensive? Is that a capability that would available to your customers? With this integration with Google? >> Yeah, so where we are going with this is we are saying, IBM DSX and other solutions that are built on top of HDP, they can transparently take advantage of the fact that they have HDP compute infrastructure to run against. So, you can run your machine learning training jobs, you can run your scoring jobs and you can have the same unmodified DSX experience whether you're running against an on-premise HDP environment or an in-cloud HDP environment. Further, that's sort of the benefit for partners and partner solutions. From a customer standpoint, the big value prop here is that customers, they're used to securing and governing their data on-prem in their particular way with HDP, with Apache Ranger, Atlas, and so forth. So, when they move to the cloud, we want this experience to be seamless from a management standpoint. So, from a data management standpoint, we want all of their learning from a security and governance perspective to apply when they are running in Google Cloud as well. So, we've had this capability on Azure and on AWS, so with this partnership, we are announcing the same type of deep integration with GCP as well. >> So Hortonworks is that one pane of glass across all your product partners for all manner of jobs. Go ahead, Rebecca. >> Well, I just wanted to ask about, we've talked about the reason, the impetus for this. With the customer, it's more familiar for customers, it offers the seamless experience, But, can you delve a little bit into the business problems that you're solving for customers here? >> A lot of times, our customers are at various points on their cloud journey, that for some of them, it's very simple, they're like there's a broom coming by and the datacenter is going away in 12 months and I need to be in the cloud. So, this is where there is a wholesale movement of infrastructure from on-premise to the cloud. Others are exploring individual business use cases. So, for example, one of our large customers, a travel partner, so they are exploring their new pricing model and they want to roll out this pricing model in the cloud. They have on-premise infrastructure, they know they have that for a while. They are spinning up new use cases in the cloud typically for reasons of agility. So, if you, typically many of our customers, they operate large, multi-tenant clusters on-prem. That's nice for, so a very scalable compute for running large jobs. But, if you want to run, for example, a new version of Spark, you have to upgrade the entire cluster before you can do that. Whereas in this sort of model, what they can say is, they can bring up a new workload and just have the specific versions and dependency that it needs, independent of all of their other infrastructure. So this gives them agility where they can move as fast as... >> Through the containerization of the Spark jobs or whatever. >> Correct, and so containerization as well as even spinning up an entire new environment. Because, in the cloud, given that you have access to elastic compute resources, they can come and go. So, your workloads are much more independent of the underlying cluster than they are on-premise. And this is where sort of the core business benefits around agility, speed of deployment, things like that come into play. >> And also, if you look at the total cost of ownership, really take an example where customers are collecting all this information through the month. And, at month end, you want to do closing of books. And so that's a great example where you want ephemeral workloads. So this is like do it once in a month, finish the books and close the books. That's a great scenario for cloud where you don't have to on-premises create an infrastructure, keep it ready. So that's one example where now, in the new partnership, you can collect all the data through the on-premises if you want throughout the month. But, move that and leverage cloud to go ahead and scale and do this workload and finish the books and all. That's one, the second example I can give is, a lot of customers collecting, like they run their e-commerce platforms and all on-premises, let's say they're running it. They can still connect all these events through HDP that may be running on-premises with Kafka and then, what you can do is, in-cloud, in GCP, you can deploy HDP, HDF, and you can use the HDF from there for real-time stream processing. So, collect all these clickstream events, use them, make decisions like, hey, which products are selling better?, should we go ahead and give?, how many people are looking at that product?, or how many people have bought it?. That kind of aggregation and real-time at scale, now you can do in-cloud and build these hybrid architectures that are there. And enable scenarios where in past, to do that kind of stuff, you would have to procure hardware, deploy hardware, all of that. Which all goes away. In-cloud, you can do that much more flexibly and just use whatever capacity you have. >> Well, you know, ephemeral workloads are at the heart of what many enterprise data scientists do. Real-world experiments, ad-hoc experiments, with certain datasets. You build a TensorFlow model or maybe a model in Caffe or whatever and you deploy it out to a cluster and so the life of a data scientist is often nothing but a stream of new tasks that are all ephemeral in their own right but are part of an ongoing experimentation program that's, you know, they're building and testing assets that may be or may not be deployed in the production applications. That's you know, so I can see a clear need for that, well, that capability of this announcement in lots of working data science shops in the business world. >> Absolutely. >> And I think coming down to, if you really look at the partnership, right. There are two or three key areas where it's going to have a huge advantage for our customers. One is analytics at-scale at a lower cost, like total cost of ownership, reducing that, running at-scale analytics. That's one of the big things. Again, as I said, the hybrid scenarios. Most customers, enterprise customers have huge deployments of infrastructure on-premises and that's not going to go away. Over a period of time, leveraging cloud is a priority for a lot of customers but they will be in these hybrid scenarios. And what this partnership allows them to do is have these scenarios that can span across cloud and on-premises infrastructure that they are building and get business value out of all of these. And then, finally, we at Google believe that the world will be more and more real-time over a period of time. Like, we already are seeing a lot of these real-time scenarios with IoT events coming in and people making real-time decisions. And this is only going to grow. And this partnership also provides the whole streaming analytics capabilities in-cloud at-scale for customers to build these hybrid plus also real-time streaming scenarios with this package. >> Well it's clear from Google what the Hortonworks partnership gives you in this competitive space, in the multi-cloud space. It gives you that ability to support hybrid cloud scenarios. You're one of the premier public cloud providers and we all know about. And clearly now that you got, you've had the Hortonworks partnership, you have that ability to support those kinds of highly hybridized deployments for your customers, many of whom I'm sure have those requirements. >> That's perfect, exactly right. >> Well a great note to end on. Thank you so much for coming on theCUBE. Sudhir, Ram, that you so much. >> Thank you, thanks a lot. >> Thank you. >> I'm Rebecca Knight for James Kobielus, we will have more tomorrow from DataWorks. We will see you tomorrow. This is theCUBE signing off. >> From sunny San Jose. >> That's right.

>> Live from San Jose, in the heart of Silicon Valley, it's theCube. Covering DataWorks Summit 2018. Brought to you by Hortonworks. >> Welcome back to day two of theCube's live coverage of DataWorks here in San Jose, California. I'm your host, Rebecca Knight along with my co-host James Kobielus. James, it's great to be here with you in the hosting seat again. >> Day two, yes. >> Exactly. So here we are, this conference, 2,100 attendees from 32 countries, 23 industries. It's a relatively big show. They do three of them during the year. One of the things that I really-- >> It's a well-established show too. I think this is like the 11th year since Yahoo started up the first Hadoop summit in 2008. >> Right, right. >> So it's an established event, yeah go. >> Exactly, exactly. But I really want to talk about Hortonworks the company. This is something that you had brought up in an analyst report before the show started and that was talking about Hortonworks' cash flow positivity for the first time. >> Which is good. >> Which is good, which is a positive sign and yet what are the prospects for this company's financial health? We're still not seeing really clear signs of robust financial growth. >> I think the signs are good for the simple reason they're making significant investments now to prepare for the future that's almost inevitable. And the future that's almost inevitable, and when I say the future, the 2020s, the decade that's coming. Most of their customers will shift more of their workloads, maybe not entirely yet, to public cloud environments for everything they're doing, AI, machine learning, deep learning. And clearly the beneficiaries of that trend will be the public cloud providers, all of whom are Hortonworks' partners and established partners, AWS, Microsoft with Azure, Google with, you know, Google Cloud Platform, IBM with IBM Cloud. Hortonworks, and this is... You know, their partnerships with these cloud providers go back several years so it's not a new initiative for them. They've seen the writing on the wall practically from the start of Hortonworks' founding in 2011 and they now need to go deeper towards making their solution portfolio capable of being deployable on-prem, in cloud, public clouds, and in various and sundry funky combinations called hybrid multi-clouds. Okay, so, they've been making those investments in those partnerships and in public cloud enabling the Hortonworks Data Platform. Here at this show, DataWorks 2018 here in San Jose, they've released the latest major version, HDP 3.0 of their core platform with a lot of significant enhancements related to things that their customers are increasingly doing-- >> Well I want to ask you about those enhancements. >> But also they have partnership announcements, the deep ones of integration and, you know, lift and shift of the Hortonworks portfolio of HDP with Hortonworks DataFlow and DataPlane Services, so that those solutions can operate transparently on those public cloud environments as the customers, as and when the customers choose to shift their workloads. 'Cause Hortonworks really... You know, like Scott Gnau yesterday, I mean just laid it on the line, they know that the more of the public cloud workloads will predominate now in this space. They're just making these speculative investments that they absolutely have to now to prepare the way. So I think this cost that they're incurring now to prepare their entire portfolio for that inevitable future is the right thing to do and that's probably why they still have not attained massive rock and rollin' positive cash flow yet but I think that they're preparing the way for them to do so in the coming decade. >> So their financial future is looking brighter and they're doing the right things. >> Yeah, yes. >> So now let's talk tech. And this is really where you want to be, Jim, I know you. >> Oh I get sleep now and I don't think about tech constantly. >> So as you've said, they're really doing a lot of emphasis now on their public cloud partnerships. >> Yes. >> But they've also launched several new products and upgrades to existing products, what are you seeing that excites you and that you think really will be potential game changers? >> You know, this is geeky but this is important 'cause it's at the very heart of Hortonworks Data Platform 3.0, containerization of more... When you're a data scientist, and you're building a machine learning model using data that's maintained, and is persisted, and processed within Hortonworks Data Platform or any other big data platform, you want the ability increasingly for developing machine learning, deep learning, AI in general, to take that application you might build while you're using TensorFlow models, that you build on HDP, they will containerize it in Docker and, you know, orchestrate it all through Kubernetes and all that wonderful stuff, and deploy it out, those AI, out to increasingly edge computing, mobile computing, embedded computing environments where, you know, the real venture capital mania's happening, things like autonomous vehicles, and you know, drones, and you name it. So the fact is that Hortonworks has made that in many ways the premier new feature of HDP 3.0 announced here this week at the show. That very much harmonizes with what their partners, where their partners are going with containerization of AI. IBM, one of their premier partners, very recently, like last month, I think it was, announced the latest version of IBM, what do they call it, IBM Cloud Private, which has embedded as a core feature containerization within that environment which is a prem-based environment of AI and so forth. The fact that Hortonworks continues to maintain close alignment with the capabilities that its public cloud partners are building to their respective portfolios is important. But also Hortonworks with its, they call it, you know, a single pane of glass, the DataPlane Services for metadata and monitoring and governance and compliance across this sprawling hybrid multi-cloud, these scenarios. The fact that they're continuing to make, in fact, really focusing on deep investments in that portfolio, so that when an IBM introduces or, AWS, whoever, introduces some new feature in their respective platforms, Hortonworks has the ability to, as it were, abstract above and beyond all of that so that the customer, the developer, and the data administrator, all they need to do, if they're a Hortonworks customer, is stay within the DataPlane Services and environment to be able to deploy with harmonized metadata and harmonized policies, and harmonized schemas and so forth and so on, and query optimization across these sprawling environments. So Hortonworks, I think, knows where their bread is buttered and it needs to stay on the DPS, DataPlane Services, side which is why a couple months ago in Berlin, Hortonworks made a, I think, the most significant announcement of the year for them and really for the industry, was that they announced the Data Steward Studio in Berlin. Tech really clearly was who addressed the GDPR mandate that was coming up but really did a stewardship as an end-to-end workflow for lots of, you know, core enterprise applications, absolutely essential. Data Steward Studio is a DataPlane Service that can operate across multi-cloud environments. Hortonworks is going to keep on, you know... They didn't have a DPS, DataPlane Services, announcements here in San Jose this week but you can best believe that next year at this time at this show, and in the interim they'll probably have a number of significant announcements to deepen that portfolio. Once again it's to grease the wheels towards a more purely public cloud future in which there will be Hortonworks DNA inside most of their customers' environments going forward. >> I want to ask you about themes of this year's conference. The thing is is that you were in Berlin at the last big Hortonworks DataWorks Summit. >> (speaks in foreign language) >> And really GDPR dominated the conversations because the new rules and regulations hadn't yet taken effect and companies were sort of bracing for what life was going to be like under GDPR. Now the rules are here, they're here to stay, and companies are really grappling with it, trying to understand the changes and how they can exist in this new regime. What would you say are the biggest themes... We're still talking about GDPR, of course, but what would you say are the bigger themes that are this week's conference? Is it scalability, is it... I mean, what would you say we're going, what do you think has dominated the conversations here? >> Well scalability is not the big theme this week though there are significant scalability announcements this week in the context of HDP 3.0, the ability to persist in a scale-out fashion across multi-cloud, billions of files. Storage efficiency is an important piece of the overall announcement with support for erasure coding, blah blah blah. That's not, you know, that's... Already, Hortonworks, like all of their cloud providers and other big data providers, provide very scalable environments for storage, workload management. That was not the hugest, buzzy theme in terms of the announcements this week. The buzz of course was HDP 3.0. Containerization, that's important, but you know, we just came out of the day two keynote. AI is not a huge focus yet for a lot of the Hortonworks customers who are here, the developers. They're, you know, most of their customers are not yet that far along in their deep learning journeys and whatever but they're definitely going there. There's plenty of really cool keynote discussions including the guy with the autonomous vehicles or whatever that, the thing we just came out of. That was not the predominant theme this week here in terms of the HDP 3.0. I think what it comes down to is that with HDP 3.0... Hive, though you tend to take it for granted, it's been in Hadoop from the very start, practically, Hive is now a full enterprise database and that's the core, one of the cores, of HDP 3.0. Hive itself, Hive 3.0 now is its version, is ACID compliant and that may be totally geeky to the most of the world but that enables it to support transactional applications. So more big data in every environment is supporting more traditional enterprise application, transactional applications that require like two-phase commit and all that goodness. The fact is, you know, Hortonworks have, from what I can see, is the first of the big data vendors to incorporate those enhancements to Hive 3.0 because they're so completely tuned in to the Hive environment in terms of a committer. I think in many ways that is the predominant theme in terms of the new stuff that will actually resonate with the developers, their customers here at the show. And with the, you know, enterprises in general, they can put more of their traditional enterprise application workloads on big data environments and specifically, Hortonworks hopes, its HDP 3.0. >> Well I'm excited to learn more here at the on theCube with you today. We've got a lot of great interviews lined up and a lot of interesting content. We got a great crew too so this is a fun show to do. >> Sure is. >> We will have more from day two of the.

>> Live from San Jose, in the heart of Silicon Valley, It's theCUBE, covering DataWorks Summit 2018. Brought to you by Hortonworks. >> Welcome back to theCUBE's live coverage of DataWorks here in San Jose, California, I'm your host, along with my cohost, James Kobielus. We're joined by Tendu Yogurtcu, she is the CTO of Syncsort. Thanks so much for coming on theCUBE, for returning to theCUBE I should say. >> Thank you Rebecca and James. It's always a pleasure to be here. >> So you've been on theCUBE before and the last time you were talking about Syncsort's growth. So can you give our viewers a company update? Where you are now? >> Absolutely, Syncsort has seen extraordinary growth within the last the last three year. We tripled our revenue, doubled our employees and expanded the product portfolio significantly. Because of this phenomenal growth that we have seen, we also embarked on a new initiative with refreshing our brand. We rebranded and this was necessitated by the fact that we have such a broad portfolio of products and we are actually showing our new brand here, articulating the value our products bring with optimizing existing infrastructure, assuring data security and availability and advancing the data by integrating into next generation analytics platforms. So it's very exciting times in terms of Syncsort's growth. >> So the last time you were on the show it was pre-GT prop PR but we were talking before the cameras were rolling and you were explaining the kinds of adoption you're seeing and what, in this new era, you're seeing from customers and hearing from customers. Can you tell our viewers a little bit about it? >> When we were discussing last time, I talked about four mega trends we are seeing and those mega trends were primarily driven by the advanced business and operation analytics. Data governance, cloud, streaming and data science, artificial intelligence. And we talked, we really made a lot of announcement and focus on the use cases around data governance. Primarily helping our customers for the GDPR Global Data Protection Regulation initiatives and how we can create that visibility in the enterprise through the data by security and lineage and delivering trust data sets. Now we are talking about cloud primarily and the keynotes, this event and our focus is around cloud, primarily driven by again the use cases, right? How the businesses are adopting to the new era. One of the challenges that we see with our enterprise customers, over 7000 customers by the way, is the ability to future-proof their applications. Because this is a very rapidly changing stack. We have seen the keynotes talking about the importance of how do you connect your existing infrastructure with the future modern, next generation platforms. How do you future-proof the platform, make a diagnostic about whether it's Amazon, Microsoft of Google Cloud. Whether it's on-premise in legacy platforms today that the data has to be available in the next generation platforms. So the challenge we are seeing is how do we keep the data fresh? How do we create that abstraction that applications are future-proofed? Because organizations, even financial services customers, banking, insurance, they now have at least one cluster running in the public cloud. And there's private implementations, hybrid becomes the new standard. So our focus and most recent announcements have been around really helping our customers with real-time resilient changes that capture, keeping the data fresh, feeding into the downstream applications with the streaming and messaging data frames, for example Kafka, Amazon Kinesis, as well as keeping the persistent stores and how to Data Lake on-premise in the cloud fresh. >> Puts you into great alignment with your partner Hortonworks so, Tendu I wonder if we are here at DataWorks, it's Hortonworks' show, if you can break out for our viewers, what is the nature, the levels of your relationship, your partnership with Hortonworks and how the Syncsort portfolio plays with HDP 3.0 with Hortonworks DataFlow and the data plan services at a high level. >> Absolutely, so we have been a longtime partner with Hortonworks and a couple of years back, we strengthened our partnership. Hortonworks is reselling Syncsort and we have actually a prescriptive solution for Hadoop and ETL onboarding in Hadoop jointly. And it's very complementary, our strategy is very complementary because what Hortonworks is trying and achieving, is creating that abstraction and future-proofing and interaction consistency around referred as this morning. Across the platform, whether it's on-premise or in the cloud or across multiple clouds. We are providing the data application layer consistency and future-proofing on top of the platform. Leveraging the tools in the platform for orchestration, integrating with HTP, certifying with Trange or HTP, all of the tools DataFlow and at last of course for lineage. >> The theme of this conference is ideas, insights and innovation and as a partner of Hortonworks, can you describe what it means for you to be at this conference? What kinds of community and deepening existing relationships, forming new ones. Can you talk about what happens here? >> This is one of the major events around data and it's DataWorks as opposed to being more specific to the Hadoop itself, right? Because stack is evolving and data challenges are evolving. For us, it means really the interactions with the customers, the organizations and the partners here. Because the dynamics of the use cases is also evolving. For example Data Lake implementations started in U.S. And we started MER European organizations moving to streaming, data streaming applications faster than U.S. >> Why is that? >> Yeah. >> Why are Europeans moving faster to streaming than we are in North America? >> I think a couple of different things might participate. The open sources really enabling organizations to move fast. When the Data Lake initiative started, we have seen a little bit slow start in Europe but more experimentation with the Open Source Stack. And by that the more transformative use cases started really evolving. Like how do I manage interactions of the users with the remote controls as they are watching live TV, type of transformative use cases became important. And as we move to the transformative use cases, streaming is also very critical because lots of data is available and being able to keep the cloud data stores as well as on-premise data stores and downstream applications with fresh data becomes important. We in fact in early June announced that Syncsort's now's a part of Microsoft One Commercial Partner Program. With that our integrate solutions with data integration and data quality are Azure gold certified and Azure ready. We are in co-sale agreement and we are helping jointly a lot of customers, moving data and workloads to Azure and keeping those data stores close to platforms in sync. >> Right. >> So lots of exciting things, I mean there's a lot happening with the application space. There's also lots still happening connected to the governance cases that we have seen. Feeding security and IT operations data into again modern day, next generation analytics platforms is key. Whether it's Splunk, whether it's Elastic, as part of the Hadoop Stack. So we are still focused on governance as part of this multi-cloud and on-premise the cloud implementations as well. We in fact launched our Ironstream for IBMI product to help customers, not just making this state available for mainframes but also from IBMI into Splunk, Elastic and other security information and event management platforms. And today we announced work flow optimization across on-premise and multi-cloud and cloud platforms. So lots of focus across to optimize, assure and integrate portfolio of products helping customers with the business use cases. That's really our focus as we innovate organically and also acquire technologies and solutions. What are the problems we are solving and how we can help our customers with the business and operation analytics, targeting those mega trends around data governance, cloud streaming and also data science. >> What is the biggest trend do you think that is sort of driving all of these changes? As you said, the data is evolving. The use cases are evolving. What is it that is keeping your customers up at night? >> Right now it's still governance, keeping them up at night, because this evolving architecture is also making governance more complex, right? If we are looking at financial services, banking, insurance, healthcare, there are lots of existing infrastructures, mission critical data stores on mainframe IBMI in addition to this gravity of data changing and lots of data with the online businesses generated in the cloud. So how to govern that also while optimizing and making those data stores available for next generation analytics, makes the governance quite complex. So that really keeps and creates a lot of opportunity for the community, right? All of us here to address those challenges. >> Because it sounds to me, I'm hearing Splunk, Advanced Machine did it, I think of the internet of things and sensor grids. I'm hearing IBM mainframes, that's transactional data, that's your customer data and so forth. It seems like much of this data that you're describing that customers are trying to cleanse and consolidate and provide strict governance on, is absolutely essential for them to drive more artificial intelligence into end applications and mobile devices that are being used to drive the customer experience. Do you see more of your customers using your tools to massage the data sets as it were than data scientists then use to build and train their models for deployment into edge applications. Is that an emerging area where your customers are deploying Syncsort? >> Thank you for asking that question. >> It's a complex question. (laughing) But thanks for impacting it... >> It is a complex question but it's very important question. Yes and in the previous discussions, we have seen, and this morning also, Rob Thomas from IBM mentioned it as well, that machine learning and artificial intelligence data science really relies on high-quality data, right? It's 1950s anonymous computer scientist says garbage in, garbage out. >> Yeah. >> When we are using artificial intelligence and machine learning, the implications, the impact of bad data multiplies. Multiplies with the training of historical data. Multiplies with the insights that we are getting out of that. So data scientists today are still spending significant time on preparing the data for the iPipeline, and the data science pipeline, that's where we shine. Because our integrate portfolio accesses the data from all enterprise data stores and cleanses and matches and prepares that in a trusted manner for use for advanced analytics with machine learning, artificial intelligence. >> Yeah 'cause the magic of machine learning for predictive analytics is that you build a statistical model based on the most valid data set for the domain of interest. If the data is junk, then you're going to be building a junk model that will not be able to do its job. So, for want of a nail, the kingdom was lost. For want of a Syncsort, (laughing) Data cleansing and you know governance tool, the whole AI superstructure will fall down. >> Yes, yes absolutely. >> Yeah, good. >> Well thank you so much Tendu for coming on theCUBE and for giving us a lot of background and information. >> Thank you for having me, thank you. >> Good to have you. >> Always a pleasure. >> I'm Rebecca Knight for James Kobielus. We will have more from theCUBE's live coverage of DataWorks 2018 just after this. (upbeat music)

>> Narrator: Live from San Francisco, it's the Cube, covering Flink Forward, brought to you by Data Artisans. (tech music) >> Hi, this is George Gilbert, we're at Flink Forward, the user conference for the Apache Flink Community, sponsored by Data Artisans. We are in San Francisco. This is the second Flink Forward conference here in San Francisco. And we have a very imminent guest, with a long pedigree, Holden Karau, formerly of IBM, and Apache Spark fame, putting Apache Spark and Python together. >> Yes. >> And now, Holden is at Google, focused on the Beam API, which is an API that makes it possible to write portable stream processing applications across Google's Dataflow, as well as Flink and other stream processors. >> Yeah. >> And Holden has been working on integrating it with the Google TensorFlow framework, also open-sourced. Yes. >> So, Holden, tell us about the objective of putting these together. What type of use cases.... >> So, I think it's really exciting. And it's still very early days, I want to be clear. If you go out there and run this code, you are going to get a lot of really weird errors, but please tell us about the errors you get. The goal is really, and we see this in Spark, with the pipeline APIs, that most of our time in machine learning is spent doing data preparation. We have to get our data in a format where we can do our machine learning on top of it. And the tricky thing about the data preparation is that we also often have to have a lot of the same preparation code available to use when we're making our predictions. And what this means is that a lot people essentially end up having to write, like, a stream-processing job to do their data preparation, and they have to write a corresponding online serving job, to do similar data preparation for when they want to make real predictions. And by integrating tf.Transform and things like this into the Beam ecosystem, the idea is that people can write their data preparation in a simple, uniform way, that can be taken from the training time into the online serving time, without them having to rewrite their code, removing the potential for mistakes where we like, change one variable slightly in one place and forget to update it in another. And just really simplifying the deployment process for these models. >> Okay, so help us tie that back to, in this case, Flink. >> Yes. >> And also to clarify, that data prep.... My impression was data prep was a different activity. It was like design time and serving was run time. But you're saying that they can be better integrated? >> So, there's different types of data prep. Some types of data prep would be things like removing invalid records. And if I'm doing that, I don't have to do that at serving time. But one of the classic examples for data prep would be tokenizing my inputs, or performing some kind of hashing transformation. And if I do that, when I get new records to predict, they won't be in a pre-tokenized form, or they won't be hashed correctly. And my model won't be able to serve on these sort of raw inputs. So I have to re-create the data prep logic that I created for training at serving time. >> So, by having common Beam API and the common provider underneath it, like Flink and TensorFlow, it's the repeatable activities for transforming data to make it ready to feed to a machine-learning model that you want those.... It would be ideal to have those transformation activities be common in your prep work, and then in the production serving. >> Yes, very true. >> So, tell us what type of customers want to write to the Beam API and have that portability? >> Yeah, so that's a really good question. So, there's a lot of people who really want portability outside of Google Cloud, and that's one group of people, essentially people who want to adopt different Google Cloud technologies, but they don't want be locked into Google Cloud forever. Which is completely understandable. There are other people who are more interested in being able to switch streaming engines, like, they want to be able to switch between Spark and Flink. And those are people who want to try out different streaming engines without having to rewrite their entire jobs. >> Does Spark Structured Streaming support the Beam API? >> So, right now, the Spark support for Beam is limited. It's in the old Dstream API, it's not on top of the Structured Streaming API. It's a thing we're actively discussing on the mailing list, how to go about doing. Because there's a lot of intricacies involved in bringing new APIs in line. And since it already works there, there's less of a pressure. But it's something that we should look at more of. Where was I going with this? So the other one that I see, is like, Flink is a wonderful API, but it's very Java-focused. And so, Java's great, everyone loves it, but a lot of cool things that are being done nowadays, are being built in Python, like TensorFlow. There's a lot of really interesting machine learning and deep learning stuff happening in Python. Beam gives a way for people to work with Python, across these different engines. Flink supports Python, but it's maybe not a first class citizen. And the Beam Python support is still a work in progress. We're working to get it to be better, but it's.... You can see the demos this afternoon, although if you're not here, you can't see the demo, but you can see the work happening in GitHub. And there's also work being done to support Go. >> In to support Go. >> Which is a little out of left field. >> So, would it be fair to say that the value of Beam, for potential Flink customers, they can work and start on Google Cloud platform. They can start on one of several stream processors. They can move to another one later, and they also inherit the better language support, or bindings from the Beam API? >> I think that's very true. The better language support, it's better for some languages, it's probably not as good for others. It's somewhat subjective, like what better language support is. But I think definitely for Go, it's pretty clear. This stuff is all stuff that's in the master branch, it's not released today. But if people are looking to play with it, I think it's really exciting. They can go and check it out from GitHub, and build it locally. >> So, what type of customers do you see who have moved into production with machine learning? >> So the.... >> And the streaming pipelines? >> The biggest customer that's in production is obviously, or not obviously, is Spotify. One of them is Spotify. They give a lot of talks about it. Because I didn't know we were going to be talking today, I didn't have a chance to go through my customer list and see who's okay with us mentioning them publicly. I'll just stick with Spotify. >> Without the names, the sort of use cases and the general industry.... >> I don't want to get in trouble. >> Okay. >> I'm just going to ... sorry. >> Okay. So then, let's talk about, does Google view Dataflow as their sort of strategic successor to map produce? >> Yes, so.... >> And is that a competitor then to Flink? >> I think Flink and Dataflow can be used in some of the same cases. But, I think they're more complementary. Flink is something you can run on-prem. You can run it in different Defenders. And Dataflow is very much like, "I can run this on Google Cloud." And part of the idea with Beam is to make it so that people who want to write Dataflow jobs but maybe want the flexibility to go back to something else later can still have that. Yeah, we couldn't swap in Flink or Dataflow execution engines if we're on Google Cloud, but.... We're not, how do I put it nicely? Provided people are running this stuff, they're burning CPU cycles, I don't really care if they're running Dataflow or Flink as the execution engine. Either way, it's a party for me, right? >> George: Okay. >> It's probably one of those, sort of, friendly competitions. Where we both push each other to do better and add more features that the respective projects have. >> Okay, 30 second question. >> Cool. >> Do you see people building stream processing applications with machine learning as part of it to extend existing apps or for ground up new apps? >> Totally. I mostly see it as extending existing apps. This is obviously, possibly a bias, just for the people that I talk to. But, going ground up with both streaming and machine learning, at the same time, like, starting both of those projects fresh is a really big hurdle to get over. >> George: For skills. >> For skills. It's really hard to pick up both of those at the same time. It's not impossible, but it's much more likely you'll build something ... maybe you'll build a batch machine learning system, realize you want to productionize your results more quickly. Or you'll build a streaming system, and then want to add some machine learning on top of it. Those are the two paths that I see. I don't see people jumping head first into both at the same time. But this could change. Batch has been King for a long time and streaming is getting it's day in the sun. So, we could start seeing people becoming more adventurous and doing both, at the same time. >> Holden, on that note, we'll have to call it a day. That was most informative. >> It's really good to see you again. >> Likewise. So this is George Gilbert. We're on the ground at Flink Forward, the Apache Flink user conference, sponsored by Data Artisans. And we will be back in a few minutes after this short break. (tech music)

>> Narrator: Live from San Jose, it's the Cube. Presenting Big Data Silicon Valley. Brought to you by SiliconANGLE Media and its ecosystem partners. >> Welcome back to the Cube's continuing coverage of Big Data SV. >> This is out tenth Big Data event, our fifth year in San Jose. We are down the street from the Strata Data Conference. We invite you to come down and join us, come on down! We are at Forager Tasting Room & Eatery, super cool place. We've got a cocktail event tonight, and a endless briefing tomorrow morning. We are excited to welcome back to the Cube, Scott Gnau, the CTO of Hortonworks. Hey, Scott, welcome back. >> Thanks for having me, and I really love what you've done with the place. I think there's as much energy here as I've seen in the entire show. So, thanks for having me over. >> Yeah! >> We have done a pretty good thing to this place that we're renting for the day. So, thanks for stopping by and talking with George and I. So, February, Hortonworks announced some news about Hortonworks DataFlow. What was in that announcement? What does that do to help customers simplify data in motion? What industries is it going to be most impactful for? I'm thinking, you know, GDPR is a couple months away, kind of what's new there? >> Well, yeah, and there are a couple of topics in there, right? So, obviously, we're very committed to, which I think is one of our unique value propositions, is we're committed to really creating an easy to use data management platform, as it were, for the entire lifecycle of data, from one data created at the edge and as data are streaming from one place to another place, and, at rest, analytics get run, analytics get pushed back out to the edge. So, that entire lifecycle is really the footprint that we're looking at, and when you dig a level into that, obviously, the data in motion piece is usually important, and So I think one a the things that we've looked at is we don't want to be just a streaming engine or just a tool for creating pipes and data flows and so on. We really want to create that entire experience around what needs to happen for data that's moving, whether it be acquisition at the edge in a protected way with provenance and encryption, whether it be applying streaming analytics as the data are flowing and everywhere kind of in between, and so that's what HDF represents, and what we released in our latest release, which, to your point, was just a few weeks ago, is a way for our customers to go build their data in motion applications using a very simple drag and drop GUI interface. So, they don't have to understand all of the different animals in the zoo, and the different technologies that are in play. It's like, "I want to do this." Okay, here's a GUI tool, you can have all of the different operators that are represented by the different underlying technologies that we provide as Hortonworks DataFlow, and you can stream them together, and then, you can make those applications and test those applications. One of the biggest enhancements that we did, is we made it very easy then for once those things are built in a laptop environment or in a dev environment, to be published out to production or to be published out to other developers who might want to enhance them and so on. So, the idea is to make it consumable inside of an enterprise, and when you think about data in motion and IOT and all those use cases, it's not going to be one department, one organization, or one person that's doing it. It's going to be a team of people that are distributed just like the data and the sensors, and, so, being able to have that sharing capability is what we've enhanced in the experience. >> So, you were just saying, before we went live, that you're here having speed dates with customers. What are some of the things... >> It's a little bit more sincere than that, but yeah. >> (laughs) Isn't speed dating sincere? It's 2018, I'm not sure. (Scott laughs) What are some of the things that you're hearing from customers, and how is that helping to drive what's coming out from Hortonworks? >> So, the two things that I'm hearing right, number one, certainly, is that they really appreciate our approach to the entire lifecycle of data, because customers are really experiencing huge data volume increases and data just from everywhere, and it's no longer just from the ERP system inside the firewall. It's from third party, it's from Sensors, it's from mobile devices, and, so, they really do appreciate kind of the territory that we cover with the tools and technologies we bring to market, and, so, that's been very rewarding. Clearly, customers who are now well into this path, they're starting to think about, in this new world, data governance, and data governance, I just took all of the energy out of the room, governance, it sounds like, you know, hard. What I mean by data governance, really, is customers need to understand, with all of this diverse, connected data everywhere, in the cloud, on PRIM, then Sensors, third party, partners, is, frankly, they need a trail of breadcrumbs that say what is it, where'd it come from, who had access to it, and then, what did they do with it? If you start to piece that together, that's what they really need to understand, the data estate that belongs to them, so they can turn that into refined product, and, so, when you then segway in one of your earlier questions, that GDPR is, certainly, a triggering point where if it's like, okay, the penalties are huge, oh my God, it's a whole new set of regulations that I have to comply with, and when you think about that trail of breadcrumbs that I just described, that actually becomes a roadmap for compliance under regulations like GDPR, where if a European customer calls up and says, "Forget my data.", the only way that you can guarantee that you forgot that person's data, is to actually understand where it all is, and that requires proper governance, tools, and techniques, and, so, when I say governance, it's, really, not like, you know, the governor and the government, and all that. That's an aspect, but the real, important part is how do I keep all of that connectivity so that I can understand the landscape of data that I've got access to, and I'm hearing a lot of energy around that, and when you think about an IOT kind of world, distributed processing, multiple hybrid cloud footprints, data is just everywhere, and, so, the perimeter is no longer fixed, it's kind of variable, and being able to keep track of that is a very important thing for our customers. >> So, continuing on that theme, Scott. Data lakes seem to be the first major new repository we added after we had data warehouses and data marts, and it looked like the governance solutions were sort of around that perimeter of the data lake. Tell us, you were alluding to, sort of, how many more repositories, whether at rest or in motion, there are for data. Do we have to solve the governance problem end-to-end before we can build meaningful applications? >> So, I would argue personally, that governance is one of the most strategic things for us as an industry, collectively, to go solve in a universal way, and what I mean by that, is throughout my career, which is probably longer than I'd like to admit, in an EDW centric world, where things are somewhat easier in terms of the perimeter and where the data came from, data sources were much more controlled, typically ERP systems, owned wholly by a company. Even in that era, true data governance, meta data management, and that provenance was never really solved adequately. There were 300 different solutions, none of which really won. They were all different, non-compatible, and the problem was easier. In this new world, with connected data, the problem is infinitely more difficult to go solve, and, so, that same kind of approach of 300 different proprietary solutions I don't think is going to work. >> So, tell us, how does that approach have to change and who can make that change? >> So, one of the things, obviously, that we're driving is we're leveraging our position in the open community to try to use the community to create that common infrastructure, common set of APIs for meta data management, and, of course, we call that Apache Atlas, and we work with a lot of partners, some of whom are customers, some of whom are other vendors, even some of whom could be considered competitors, to try to drive an Apache open source kind of project to become that standard layer that's common into which vendors can bring their applications. So, now, if I have a common API for tracking meta data in that trail of breadcrumbs that's commonly understood, I can bring in an application that helps customers go develop the taxonomy of the rules that they want to implement, and, then, that helps visualize all of the other functionality, which is also extremely important, and that's where I think specialization comes into play, but having that common infrastructure, I think, is a really important thing, because that's going to enable data, data lakes, IOT to be trusted, and if it's not trusted, it's not going to be successful. >> Okay, there's a chicken and an egg there it sounds like, potentially. >> Am I the chicken or the egg? >> Well, you're the CTO. (Lisa laughs) >> Okay. >> The thing I was thinking of was, the broader the scope of trust that you're trying to achieve at first, the more difficult the problem, do you see customers wanting to pick off one high value application, not necessarily that's about managing what's in Atlas, in the meta data, so much as they want to do an IOT app and they'll implement some amount of governance to solve that app. In other words, which comes first? Do they have to do the end-to-end meta data management and governance, or do they pick a problem off first? >> In this case, I think it's chicken or egg. I mean, you could start from either point. I see customers who are implementing applications in the IOT space, and they're saying, "Hey, this requires a new way to think of governance, "so, I'm going to go and build that out, but I'm going to "think about it being pluggable into the next app." I also see a lot of customers, especially in highly regulated industries, and especially in highly regulated jurisdictions, who are stepping back and saying, "Forget the applications, this is a data opportunity, "and, so, I want to go solve my data fabric, "and I want to have some consistency across "that data fabric into which I can publish data "for specific applications and guarantee "that, wholistically, I am compliant "and that I'm sitting inside of our corporate mission "and all of those things." >> George: Okay. >> So, one of the things you mention, and we talk about this a lot, is the proliferation of data. It's so many, so many different sources, and companies have an opportunity, you had mentioned the phrase data opportunity, there is massive opportunity there, but you said, you know, from even a GDR perspective alone, I can't remove the data if I don't know where it is to the breadcrumbs. As a marketer, we use terms like get a 360 degree view of your customer. Is that actually really something that customers can achieve leveraging a data. Can they actually really get, say a retailer, a 360, a complete view of their customer? >> Alright, 358. >> That's pretty good! >> And we're getting there. (Lisa laughs) Yeah, I mean, obviously, the idea is to get a much broader view, and 360 is a marketing term. I'm not a marketing person, >> Yes. But it, certainly, creates a much broader view of highly personalized information that help you interact with your customer better, and, yes, we're seeing customers do that today and have great success with it and actually change and build new business models based on that capability, for sure. The folks who've done that have realized that in this new world, the way that that works is you have to have a lot of people have access to a lot of data, and that's scary, because that's not the way it used to be, right? >> Right. >> It used to be you go to the DBA and you ask for access, and then, your boss has to sign off and say it's what you asked for. In this world, you need to have access to all of it. So, when you think about this new governance capability where as part of the governance integrated with security, personalized information can be encrypted, it can be blurred out, but you still have access to the data to look at the relationships to be found in the data to build out those sophisticated models. So, that's where not only is it a new opportunity for governance just because the sources, the variety at the different landscape, but it's, ultimately, very much required, because if you're the CSO, you're not going to give access to the marketing team all of its customer data unless you understand that, right, but it has to be, "I'm just giving it to you, "and I know that it's automatically protected." versus, "I'm going to let you ask for it." to be successful. >> Right. >> I guess, following up on that, it sounds like what we were talking about, chicken or egg. Are you seeing an accelerating shift from where data is sort of collected, centrally, from applications, or, what we hear on Amazon, is the amount coming off the edge is accelerating. >> It is, and I think that that is a big drive to, frankly, faster clouded option, you know, the analytic space, particularly, has been a laggard in clouded option for many reasons, and we've talked about it previously, but one of the biggest reasons, obviously, is that data has gravity, data movement is expensive, and, so, now, when you think about where data is being created, where it lives, being further out on the edge, and may live its entire lifecycle in the cloud, you're seeing a reversal of gravity more towards cloud, and that, again, creates more opportunities in terms of driving a more varied perimeter and just keeping track of where all the assets are. Finally, I think it also leads to this notion of managing entire lifecycle of data. One of the implications of that is if data is not going to be centralized, it's going to live in different places, applications have to be portable to move to where the data exists. So, when I think about that landscape of creating ubiquitous data management within Hortonworks' portfolio, that's one of the big values that we can create for our customers. Not only can we be an on-ramp to their hybrid architecture, but as we become that on-ramp, we can also guarantee the portability of the applications that they've built out to those cloud footprints and, ultimately, even out to the edge. >> So, a quick question, then, to clarify on that, or drill down, would that mean you could see scenarios where Hortonworks is managing the distribution of models that do the inferencing on the edge, and you're collecting, bringing back the relevant data, however that's defined, to do the retraining of any models or recreation of new models. >> Absolutely, absolutely. That's one of the key things about the NiFi project in general and Hortonworks DataFlow, specifically, is the ability to selectively move data, and the selectivity can be based on analytic models as well. So, the easiest case to think about is self-driving cars. We all understand how that works, right? A self-driving car has cameras, and it's looking at things going on. It's making decisions, locally, based on models that have been delivered, and they have to be done locally, because of latency, right, but, selectively, hey, here's something that I saw as an image I didn't recognize. I need to send that up, so that it can be added to my lexicon of what images are and what action should be taken. So, of course, that's all very futuristic, but we understand how that works, but that has application in things that are very relevant today. Think about jet engines that have diagnostics running. Do I need to send that terabyte of data an hour over an expensive thing? No, but I have a model that runs locally that says, "Wow, this thing looks interesting. "Let me send a gigabyte now for immediate action." So, that decision making capability is extremely important. >> Well, Scott, thanks so much for taking some time to come chat with us once again on the Cube. We appreciate your insights. >> Appreciate it, time flies. This is great. >> Doesn't it? When you're having fun! >> Yeah. >> Alright, we want to thank you for watching the Cube. I'm Lisa Martin with George Gilbert. We are live at Forager Tasting Room in downtown San Jose at our own event, Big Data SV. We'd love for you to come on down and join us tonight, today, tonight, and tomorrow. Stick around, we'll be right back with our next guest after a short break. (techno music) >> Narrator: Since the dawn of the cloud, the Cube

>> Welcome everybody, we're at the Flink Forward conference in San Francisco, at the Kabuki Hotel. Flink Forward U.S. is the first U.S. user conference for the Flink community sponsored by data Artisans, the creators of Flink, and we're here with special guest Kenneth Knowles-- >> Hi. >> Who works for Google and who heads up the Apache Beam Team where, just to set context, Beam is the API Or STK on which developers can build stream processing apps that can be supported by Google's Dataflow, Apache Flink, Spark, Apex, among other future products that'll come along. Ken, why don't you tell us, what was the genesis of Beam, and why did Google open up sort of the API to it. >> So, I can speak as an Apache Beam Team PMC member, that the genesis came from a combined code donation to Apache from Google Cloud Dataflow STK and there was also already written by data Artisans a Flink runner for that, which already included some portability hooks, and then there was also a runner for Spark that was written by some folks at PayPal. And so, sort of those three efforts pointed out that it was a good time to have a unified model for these DAG-based computational... I guess it's a DAG-based computational model. >> Okay, so I want to pause you for a moment. >> Yeah. >> And generally, we try to avoid being rude and cutting off our guests but, in this case, help us understand what a DAG is, and why it's so important. >> Okay, so a DAG is a directed acyclic graph, and, in some sense, if you draw a boxes and arrows diagram of your computation where you say "I read some data from here," and it goes through some filters and then I do a join and then I write it somewhere. These all end up looking what they call the DAG just because of the fact that it is the structure, and all computation sort of can be modeled this way, and in particular, these massively parallel computations profit a lot from being modeled this way as opposed to MapReduce because the fact that you have access to the entire DAG means you can perform transformations and optimizations and you have more opportunities for executing it in different ways. >> Oh, in other words, because you can see the big picture you can find, like, the shortest path as opposed to I've got to do this step, I've got to do this step and this step. >> Yeah, it's exactly like that, you're not constrained to sort of, the person writing the program knows what it is that they want to compute, and then, you know, you have very smart people writing the optimizer and the execution engine. So it may execute an entirely different way, so for example, if you're doing a summation, right, rather than shuffling all your data to one place and summing there, maybe you do some partial summations, and then you just shuffle accumulators to one place, and finish the summation, right? >> Okay, now let me bump you up a couple levels >> Yeah. >> And tell us, so, MapReduce was a trees within the forest approach, you know, lots of seeing just what's a couple feet ahead of you. And now we have the big picture that allows you to find the best path, perhaps, one way of saying it. Tell us though, with Google or with others who are using Beam-compatible applications, what new class of solutions can they build that you wouldn't have done with MapReduce before? >> Well, I guess there's... There's two main aspects to Beam that I would emphasize, there's the portability, so you can write this application without having to commit to which backend you're going to run it on. And there's... There's also the unification of streaming and batch which is not present in a number of backends, and Beam as this layer sort of makes it very easy to use sort of batch-style computation and streaming-style computation in the same pipeline. And actually I said there was two things, the third thing that actually really opens things up is that Beam is not just a portability layer across backends, it's also a portability layer across languages, so, something that really only has preliminary support on a lot of systems is Python, so, for example, Beam has a Python STK where you write a DAG description of your computation in Python, and via Beam's portability API's, one of these sort of usually Java-centric engines would be able to run that Python pipeline. >> Okay, so-- >> So, did I answer your question? >> Yes, yes, but let's go one level deeper, which is, if MapReduce, if its sweet spot was web crawl indexing in batch mode, what are some of the things that are now possible with a Beam-style platform that supports Beam, you know, underneath it, that can do this direct acyclic graph processing? >> I guess what I, I'm still learning all the different things that you can do with this style of computation, and the truth is it's just extremely general, right? You can set up a DAG, and there's a lot of talks here at Flink Forward about using a stream processor to do high frequency trading or fraud detection. And those are completely different even though they're in the same model of computation as, you know, you would still use it for things like crawling the web and doing PageRank over. Actually, at the moment we don't have iterative computations so we wouldn't do PageRank today. >> So, is it considered a complete replacement, and then new used cases for older style frameworks like MapReduce, or is it a complement for things where you want to do more with data in motion or lower latency? >> It is absolutely intended as a full replacement for MapReduce, yes, like, if you're thinking about writing a MapReduce pipeline, instead you should write a Beam pipeline, and then you should benchmark it on different Beam backends, right? >> And, so, working with Spark, working with Flink, how are they, in terms of implementing the full richness of the Beam-interface relative to the Google product Dataflow, from which I assumed Beam was derived? >> So, all of the different backends exist in sort of different states as far as implementing the full model. One thing I really want to emphasize is that Beam is not trying to take the intersection on all of these, right? And I think that your question already shows that you know this, we keep sort of a matrix on our website where we say, "Okay there's all these different "features you might want, "and then there's all these backends "you might want to run it on," and it's sort of there's can you do it, can you do it sometimes, and notes about that, we want this whole matrix to be, yes, you can use all of the model on Flink, all of it on Spark, all of it on Google Cloud Dataflow, but so they all have some gaps and I guess, yeah, we're really welcoming contributors in that space. >> So, for someone whose been around for a long time, you might think of it as an ODBC driver, where the capabilities of the databases behind it are different, and so the drivers can only support some subset of a full capability. >> Yeah, I think that there's, so, I'm not familiar enough with ODBC to say absolutely yes, absolutely no, but yes, it's that sort of a thing, it's like the JVM has many languages on it and ODBC provides this generic database abstraction. >> Is Google's goal with Beam API to make it so that customers demand a level of portability that goes not just for the on-prim products but for products that are in other public clouds, and sort of pry open the API lock in? >> So, I can't say what Google's goals are, but I can certainly say that Beam's goals are that nobody's going to be locked into a particular backend. >> Okay. >> I mean, I can't even say what Beam's goals are, sorry, those are my goals, I can speak for myself. >> Is Beam seeing so far adoption by the sort of big consumer internet companies, or has it started to spread to mainstream enterprises, or is still a little immature? >> I think Beam's still a little bit less mature than that, we're heading into our first stable release, so, we began incubating it as an Apache project about a year ago, and then, around the beginning of the new year, actually right at the end of 2016, we graduated to be an Apache top level project, so right now we're sort of on the road from we've become a top level project, we're seeing contributions ramp up dramatically, and we're aiming for a stable release as soon as possible, our next release we expect to be a stable API that we would encourage users and enterprises to adopt I think. >> Okay, and that's when we would see it in production form on the Google Cloud platform? >> Well, so the thing is that the code and the backends behind it are all very mature, but, right now, we're still sort of like, I don't know how to say it, we're polishing the edges, right, it's still got a lot of rough edges and you might encounter them if you're trying it out right now and things might change out from under you before we make our stable release. >> Understood. >> Yep. All right. Kenneth, thank you for joining us, and for the update on the Beam project and we'll be looking for that and seeing its progress over the next few months. >> Great. Thanks for having me. >> With that, I'm George Gilbert, I'm with Kenneth Knowles, we're at the dataArtisan's Flink Forward user conference in San Francisco at the Kabuki Hotel and we'll be back after a few minutes.

>> Announcer: Coverage DataWorks Summit Europe 2017 brought to you by Hortonworks. >> Welcome back everyone. Live here in Munich, Germany for theCUBE'S special presentation of Hortonworks Hadoop Summit now called DataWorks 2017. I'm John Furrier, my co-host Dave Vellante, our next guest is Shaun Connolly, Vice President of Corporate Strategy, Chief Strategy Officer. Shaun great to see you again. >> Thanks for having me guys. Always a pleasure. >> Super exciting. Obviously we always pontificating on the status of Hadoop and Hadoop is dead, long live Hadoop, but runs in demise is greatly over-exaggerated, but reality is is that no major shifts in the trends other than the fact that the amplification with AI and machine learning has upleveled the narrative to mainstream around data, big data has been written on on gen one on Hadoop, DevOps, culture, open-source. Starting with Hadoop you guys certainly have been way out in front of all the trends. How you guys have been rolling out the products. But it's now with IoT and AI as that sizzle, the future self driving cars, smart cities, you're starting to really see demand for comprehensive solutions that involve data-centric thinking. Okay, said one. Two, open-source continues to dominate MuleSoft went public, you guys went public years ago, Cloudera filed their S-1. A crop of public companies that are open-source, haven't seen that since Red Hat. >> Exactly. 99 is when Red Hat went public. >> Data-centric, big megatrend with open-source powering it, you couldn't be happier for the stars lining up. >> Yeah, well we definitely placed our bets on that. We went public in 2014 and it's nice to see that graduating class of Taal and MuleSoft, Cloudera coming out. That just I think helps socializes movement that enterprise open-source, whether it's for on-prem or powering cloud solutions pushed out to the edge, and technologies that are relevant in IoT. That's the wave. We had a panel earlier today where Dahl Jeppe from Centric of British Gas, was talking about his ... The digitization of energy and virtual power plant notions. He can't achieve that without open-source powering and fueling that. >> And the thing about it is is just kind of ... For me personally being my age in this generation of computer industry since I was 19, to see the open-source go mainstream the way it is, is even gets better every time, but it really is the thousandth flower bloom strategy. Throwing the seeds out there of innovation. I want to ask you as a strategy question, you guys from a performance standpoint, I would say kind of got hammered in the public market. Cloudera's valuation privately is 4.1 billion, you guys are close to 700 million. Certainly Cloudera's going to get a haircut looks like. The public market is based on the multiples from Dave and I's intro, but there's so much value being created. Where's the value for you guys as you look at the horizon? You're talking about white spaces that are really developing with use cases that are creating value. The practitioners in the field creating value, real value for customers. >> So you covered some of the trends, but I'll translate em into how the customers are deploying. Cloud computing and IoT are somewhat related. One is a centralization, the other is decentralization, so it actually calls for a connected data architecture as we refer to it. We're working with a variety of IoT-related use cases. Coca-Cola, East Japan spoke at Tokyo Summit about beverage replenishment analytics. Getting vending machine analytics from vending machines even on Mount Fuji. And optimizing their flow-through of inventory in just-in-time delivery. That's an IoT-related to run on Azure. It's a cloud-related story and it's a big data analytics story that's actually driving better margins for the business and actually better revenues cuz they're getting the inventory where it needs to be so people can buy it. Those are really interesting use cases that we're seeing being deployed and it's at this convergence of IoT cloud and big data. Ultimately that leads to AI, but I think that's what we're seeing the rise of. >> Can you help us understand that sort of value chain. You've got the edge, you got the cloud, you need something in-between, you're calling it connected data platform. How do you guys participate in that value chain? >> When we went public our primary workhorse platform was Hortonworks Data Platform. We had first class cloud services with Azure HDInsight and Hortonworks Data Cloud for AWS, curated cloud services pay-as-you-go, and Hortonworks DataFlow, I call as our connective tissue, it manages all of your data motion, it's a data logistics platform, it's like FedEx for data delivery. It goes all the way out to the edge. There's a little component called Minify, mini and ify, which does secure intelligent analytics at the edge and transmission. These smart manufacturing lines, you're gathering the data, you're doing analytics on the manufacturing lines, and then you're bringing the historical stuff into the data center where you can do historical analytics across manufacturing lines. Those are the use cases that are connect the data archives-- >> Dave: A subset of that data comes back, right? >> A subset of the data, yep. The key events of that data it may not be full of-- >> 10%, half, 90%? >> It depends if you have operational events that you want to store, sometimes you may want to bring full fidelity of that data so you can do ... As you manufacture stuff and when it got deployed and you're seeing issues in the field, like Western Digital Hard Drives, that failure's in the field, they want that data full fidelity to connect the data architecture and analytics around that data. You need to ... One of the terms I use is in the new world, you need to play it where it lies. If it's out at the edge, you need to play it there. If it makes a stop in the cloud, you need to play it there. If it comes into the data center, you also need to play it there. >> So a couple years ago, you and I were doing a panel at our Big Data NYC event and I used the term "profitless prosperity," I got the hairy eyeball from you, but nonetheless, we talked about you guys as a steward of the industry, you have to invest in open-source projects. And it's expensive. I mean HDFS itself, YARN, Tez, you guys lead a lot of those initiatives. >> Shaun: With the community, yeah, but we-- >> With the community yeah, but you provided contributions and co-leadership let's say. You're there at the front of the pack. How do we project it forward without making forward-looking statements, but how does this industry become a cashflow positive industry? >> Public companies since end of 2014, the markets turned beginning at 2016 towards, prior to that high growth with some losses was palatable, losses were not palatable. That his us, Splunk, Tableau most of the IT sector. That's just the nature of the public markets. As more public open-source, data-driven companies will come in I think it will better educate the market of the value. There's only so much I can do to control the stock price. What I can from a business perspective is hit key measures from a path to profitability. The end of Q4 2016, we hit what we call the just-to-even or breakeven, which is a stepping stone. On our earnings call at the end of 2016 we ended with 185 million in revenue for the year. Only five years into this journey, so that's a hard revenue growth pace and we basically stated in Q3 or Q4 of 17, we will hit operating cashflow neutrality. So we are operating business-- >> John: But you guys also hit a 100 million at record pace too, I believe. >> Yeah, in four years. So revenue is one thing, but operating margins, like if you look at our margins on our subscription business for instance, we've got 84% margin on that. It's a really nice margin business. We can make that better margins, but that's a software margin. >> You know what's ironic, we were talking about Red Hat off camera. Here's Red Hat kicking butt, really hitting all cylinders, three billion dollars in bookings, one would think, okay hey I can maybe project forth some of these open-source companies. Maybe the flip side of this, oh wow we want it now. To your point, the market kind of flipped, but you would think that Red Hat is an indicator of how an open-source model can work. >> By the way Red Hat went public in 99, so it was a different trajectory, like you know I charted their trajectory out. Oracle's trajectory was different. They didn't even in inflation adjusted dollars they didn't hit a 100 million in four years, I think it was seven or eight years or what have you. Salesforce did it in five. So these SaaS models and these subscription models and the cloud services, which is an area that's near and dear to my heart. >> John: Goes faster. >> You get multiple revenue streams across different products. We're a multi-products cloud service company. Not just a single platform. >> So we were actually teasing this out on our-- >> And that's how you grow the business, and that's how Red Hat did it. >> Well I want to get your thoughts on this while we're just kind of ripping live here because Dave and I were talking on our intro segment about the business model and how there's some camouflage out there, at least from my standpoint. One of the main areas that I was kind of pointing at and trying to poke at and want to get your reaction to is in the classic enterprise go-to-market, you have sales force expansive, you guys pay handsomely for that today. Incubating that market, getting the profitability for it is a good thing, but there's also channels, VARs, ISVs, and so on. You guys have an open-source channel that kind of not as a VAR or an ISV, these are entrepreneurs and or businesses themselves. There's got to be a monetization shift there for you guys in the subscription business certainly. When you look at these partners, they're co-developing, they're in open-source, you can almost see the dots connecting. Is this new ecosystem, there's always been an ecosystem, but now that you have kind of a monetization inherently in a pure open distribution model. >> It forces you to collaborate. IBM was on stage talking about our system certified on the Power Systems. Many may look at IBM as competitive, we view them as a partner. Amazon, some may view them as a competitor with us, they've been a great partner in our for AWS. So it forces you to think about how do you collaborate around deeply engineered systems and value and we get great revenue streams that are pulled through that they can sell into the market to their ecosystems. >> How do you vision monetizing the partners? Let's just say Dave and I start this epic idea and we create some connective tissue with your orchestrator called the Data Platform you have and we start making some serious bang. We make a billion dollars. Do you get paid on that if it's open-source? I mean would we be more subscriptions? I'm trying to see how the tide comes in, whose boats float on the rising tide of the innovation in these white spaces. >> Platform thinking is you provide the platform. You provide the platform for 10x value that rides atop that platform. That's how the model works. So if you're riding atop the platform, I expect you and that ecosystem to drive at least 10x above and beyond what I would make as a platform provider in that space. >> So you expect some contributions? >> That's how it works. You need a thousand flowers to be running on the platform. >> You saw that with VMware. They hit 10x and ultimately got to 15 or 16, 17x. >> Shaun: Exactly. >> I think they don't talk about it anymore. I think it's probably trading the other way. >> You know my days at JBoss Red Hat it was somewhere between 15 to 20x. That was the value that was created on top of the platforms. >> What about the ... I want to ask you about the forking of the Hadoop distros. I mean there was a time when everybody was announcing Hadoop distros. John Furrier announced SiliconANGLE was announcing Hadoop distro. So we saw consolidation, and then you guys announced the ODP, then the ODPI initiative, but there seems to be a bit of a forking in Hadoop distros. Is that a fair statement? Unfair? >> I think if you look at how the Linux market played out. You have clearly Red Hat, you had Conicho Ubuntu, you had SUSE. You're always going to have curated platforms for different purposes. We have a strong opinion and a strong focus in the area of IoT, fast analytic data from the edge, and a centralized platform with HDP in the cloud and on-prem. Others in the market Cloudera is running sort of a different play where they're curating different elements and investing in different elements. Doesn't make either one bad or good, we are just going after the markets slightly differently. The other point I'll make there is in 2014 if you looked at the then chart diagrams, there was a lot of overlap. Now if you draw the areas of focus, there's a lot of white space that we're going after that they aren't going after, and they're going after other places and other new vendors are going after others. With the market dynamics of IoT, cloud and AI, you're going to see folks chase the market opportunities. >> Is that dispersity not a problem for customers now or is it challenging? >> There has to be a core level of interoperability and that's one of the reasons why we're collaborating with folks in the ODPI, as an example. There's still when it comes to some of the core components, there has to be a level of predictability, because if you're an ISV riding atop, you're slowed down by death by infinite certification and choices. So ultimately it has to come down to just a much more sane approach to what you can rely on. >> When you guys announced ODP, then ODPI, the extension, Mike Olson wrote a blog saying it's not necessary, people came out against it. Now we're three years in looking back. Was he right or not? >> I think ODPI take away this year, there's more than we can do above and beyond the Hadoop platform. It's expanded to include SQL and other things recently, so there's been some movement on this spec, but frankly you talk to John Mertic at ODPI, you talk to SAS and others, I think we want to be a bit more aggressive in the areas that we go after and try and drive there from a standardization perspective. >> We had Wei Wang on earlier-- >> Shaun: There's more we can do and there's more we should do. >> We had Wei on with Microsoft at our Big Data SV event a couple weeks ago. Talk about the Microsoft relationship with you guys. It seems to be doing very well. Comments on that. >> Microsoft was one of the two companies we chose to partner with early on, so and 2011, 2012 Microsoft and Teradata were the two. Microsoft was how do I democratize and make this technology easy for people. That's manifest itself as Azure Cloud Service, Azure HDInsight-- >> Which is growing like crazy. >> Which is globally deployed and we just had another update. It's fundamentally changed our engineering and delivering model. This latest release was a cloud first delivery model, so one of the things that we're proud of is the interactive SQL and the LLAP technology that's in HDP, that went out through Azure HDInsight what works data cloud first. Then it certified in HDP 2.6 and it went power at the same time. It's that cadence of delivery and cloud first delivery model. We couldn't do it without a partnership with Microsoft. I think we've really learned what it takes-- >> If you look at Microsoft at that time. I remember interviewing you on theCUBE. Microsoft was trading something like $26 a share at that time, around their low point. Now the stock is performing really well. Stockinnetel very cloud oriented-- >> Shaun: They're very open-source. >> They're very open-source and friendly they've been donating a lot to the OCP, to the data center piece. Extremely different Microsoft, so you slipped into that beautiful spot, reacted on that growth. >> I think as one of the stalwarts of enterprise software providers, I think they've done a really great job of bending the curve towards cloud and still having a mixed portfolio, but in sending a field, and sending a channel, and selling cloud and growing that revenue stream, that's nontrivial, that's hard. >> They know the enterprise sales motions too. I want to ask you how that's going over all within Hortonworks. What are some of the conversations that you're involved in with customers today? Again we were saying in our opening segment, it's on YouTube if you're not watching, but the customers is the forcing function right now. They're really putting the pressure one the suppliers, you're one of them, to get tight, reduce friction, lower costs of ownership, get into the cloud, flywheel. And so you see a lot-- >> I'll throw in another aspect some of the more late majority adopters traditionally, over and over right here by 2025 they want to power down the data center and have more things running in the public cloud, if not most everything. That's another eight years or what have you, so it's still a journey, but this journey to making that an imperative because of the operational, because of the agility, because of better predictability, ease of use. That's fundamental. >> As you get into the connected tissue, I love that example, with Kubernetes containers, you've got developers, a big open-source participant and you got all the stuff you have, you just start to see some coalescing around the cloud native. How do you guys look at that conversation? >> I view container platforms, whether they're container services that are running one on cloud or what have you, as the new lightweight rail that everything will ride atop. The cloud currently plays a key role in that, I think that's going to be the defacto way. In particularly if you go cloud first models, particularly for delivery. You need that packaging notion and you need the agility of updates that that's going to provide. I think Red Hat as a partner has been doing great things on hardening that, making it secure. There's others in the ecosystem as well as the cloud providers. All three cloud providers actually are investing in it. >> John: So it's good for your business? >> It removes friction of deployment ... And I ride atop that new rail. It can't get here soon enough from my perspective. >> So I want to ask about clouds. You were talking about the Microsoft shift, personally I think Microsoft realized holy cow, we could actaully make a lot of money if we're selling hardware services. We can make more money if we're selling the full stack. It was sort of an epiphany and so Amazon seems to be doing the same thing. You mentioned earlier you know Amazon is a great partner, even though a lot of people look at them as a competitor, it seems like Amazon, Azure etc., they're building out their own big data stack and offering it as a service. People say that's a threat to you guys, is it a threat or is it a tailwind, is it it is what it is? >> This is why I bring up industry-wide we always have waves of centralization, decentralization. They're playing out simultaneously right now with cloud and IoT. The fact of the matter is that you're going to have multiple clouds on-prem data and data at the edge. That's the problem I am looking to facilitate and solve. I don't view them as competitors, I view them as partners because we need to collaborate because there's a value chain of the flow of the data and some of it's going to be running through and on those platforms. >> The cloud's not going to solve the edge problem. Too expensive. It's just physics. >> So I think that's where things need to go. I think that's why we talk about this notion of connected data. I don't talk hybrid cloud computing, that's for compute. I talk about how do you connect to your data, how do you know where your data is and are you getting the right value out of the data by playing it where it lies. >> I think IoT has been a great sweet trend for the big data industry. It really accelerates the value proposition of the cloud too because now you have a connected network, you can have your cake and eat it too. Central and distributed. >> There's different dynamics in the US versus Europe, as an example. US definitely we're seeing a cloud adoption that's independent of IoT. Here in Europe, I would argue the smart mobility initiatives, the smart manufacturing initiatives, and the connected grid initiatives are bringing cloud in, so it's IoT and cloud and that's opening up the cloud opportunity here. >> Interesting. So on a prospects for Hortonworks cashflow positive Q4 you guys have made a public statement, any other thoughts you want to share. >> Just continue to grow the business, focus on these customer use cases, get them to talk about them at things like DataWorks Summit, and then the more the merrier, the more data-oriented open-source driven companies that can graduate in the public markets, I think is awesome. I think it will just help the industry. >> Operating in the open, with full transparency-- >> Shaun: On the business and the code. (laughter) >> Welcome to the party baby. This is theCUBE here at DataWorks 2017 in Munich, Germany. Live coverage, I'm John Furrier with Dave Vellante. Stay with us. More great coverage coming after this short break. (upbeat music)

>> Announcer: Live from Boston, Massachusetts. This is "The Cube" covering Spark Summit East 2017. Brought to you by Databricks. Now here are your hosts Dave Volante and George Gilbert. >> Welcome back to snowy Boston everybody. This is "The Cube." The leader in live tech coverage. This is Spark Summit. Spark Summit East #SparkSummit. Bryan Duxbury's here. He's the vice president of engineering at StreamSets. Cleveland boy! Welcome to "The Cube." >> Thanks for having me. >> You've very welcome. Tell us, let's start with StreamSets. We're going to talk about Spark and some of the use cases that it's enabling and some of the integrations you're doing. But what does StreamSets do? >> Sure, StreamSets is a data movement software. So I like to think of it either the first mile or the last mile of a lot of different analytical or data movement workflows. Basically we build a product that allows you to build a workflow, or build a data pipeline that doesn't require you to code. It's a graphical user interphase for dropping an origin, several destinations, and then lightweight transformations onto a canvas. You click play and it runs. So this is kind of different than, a lot of the market today is a programming tool or a command line tool. That still requires your systems engineers or your unfortunate data scientists pretending to be systems engineers to do systems engineering. To do a science project to figure out how to move data. The challenge of data movement I think is often underplayed how challenging it is. But it's extremely tedious work. You know, you have to connect to dozens or hundreds of different data sources. Totally different schemas. Different database drivers, or systems altogether. And it break all the time. So the home-built stuff is really challenging to keep online. When it goes down, your business is not, you're not moving data. You can't actually get the insights you built in the first place. >> I remember I broke into this industry you know, in the days of mainframe. You used to read about them and they had this high-speed data mover. And it was this key component. And it had to be integrated. It had to be able to move, back then, it was large amounts of data fast. Today especially with the advent of Hadoop, people say okay don't move the data, keep it in place. Now that's not always practical. So talk about the sort of business case for starting a company that basically moves data. >> We handle basically the one step before. I agree with you completely. Many data analytical situations today where you're doing like the true, like business-oriented detail, where you're actually analyzing data and producing value, you can do it in place. Which is to say in your cluster, in your Spark cluster, all the different environments you can imagine. The problem is that if it's not there already, then it's a pretty monumental effort to get it there. I think we see. You know a lot of people think oh I can just write a SQL script, right? And that works for the first two to 20 tables you want to deploy. But for instance, in my background, I used to work at Square. I ran a data platform there. We had 500 tables we had to move on a regular basis. Coupled with a whole variety of other data sources. So at some point it becomes really impractical to hand-code these solutions. And even when you build your own framework, and you start to build tools internally, you know, it's not your job really, these companies, to build a world class data movement tool. It's their job to make the data valuable, right? And actually data movement is like utility, right. Providing the utility, really the thing to do is be productive and cost effective, right? So the reason why we build StreamSets, the reason why this thing is a thing in the first place, is because we think people shouldn't be in the business of building data movement tools. They should be in the business of moving their data and then getting on with it. Does that make sense? >> Yeah absolutely. So talk about how it all fits in with Spark generally and specifically Spark coming to the enterprise. >> Well in terms of how StreamSets connects to stuff, we deploy in every way you can imagine, whether you want to run your own premise, on your own machines, or in the Cloud. It's up to you to deploy however you like. We're not prescriptive about that. We often get deployed on the edge of clusters, wether it's your Hadoop cluster or your Spark cluster. And basically we try not to get in the way of these analysis tools. There are many great analytical tools out there like Spark is a great example. We focus really on the moving of data. So what you'll see is someone will build a Spark streaming application or some big Spark SQL thing that actually produces the reports. And we plug in ahead of that. So if you're data is being collected from, you know, Edge web logs or some thing or some Kafka thing or a third party AVI or scripting website. We do the first collection. And then it's usually picked up from there with the next tool. Whether it's Spark or other things. I'm trying to think about the right way to put this. I think that people who write Spark they should focus on the part that's like the business value for them. They should be doing the thing that actually is applying the machine learning model, or is producing the report that the CEO or CTO wants to see. And move away from the ingest part of the business. Does that make sense? >> [] Yeah. >> Yeah. When the Spark guys sort of aspire to that by saying you don't have to worry about exactly when's delivery. And you know you can make sure this sort of guarantee, you've got guarantees that will get from point A to point B. >> Bryan: Yeah. >> Things like that. But all those sources of data and all those targets, writing all those adapters is, I mean, that's been a La Brea tar pit for many companies over time. >> In essence that is our business. I think that you touch on a good point. Spark can actually do some of these things right. There's not complete, but significant overlap in some cases. But the important difference is that Spark is a cluster tool for working with cluster data. And we're not going to beat you running a Spark application for consuming from Kafka to do your analysis. But you want to use Spark for reading local files? Do you want to use Spark for reading from a mainframe? Like these are things that StreamSets is built for. And that library of connectors you're talking about, it's our bread and butter. It's not your job as a data scientist, you know, applying Spark, to build a library of connectors. So actually the challenge is not the difficulty of building any one connector, because we have that down to an art now. But we can afford to invest, we can build a portfolio of connectors. But you as a user of Spark, can only afford to do it on demand. Reactive. And so that turn around time, of the cost it might take you to build that connector is pretty significant. And actually I often see the flow side. This is a problem I faced at Square, which was that people asked me to integrate new data sources, I had to say no. Because it was too rare, it was too unusual for what we had to do. We had other things to support. So the problem with that is that I have no idea what kind of opportunity cost I left behind. Like what kind of data we didn't get, kind of analysis we couldn't do. And with an approach like StreamSets, you can solve that problem sort of up front even. >> So sort of two follow ups. One is it would seem to be an evergreen effort to maintain the existing connectors. >> Bryan: Certainly. >> And two, is there a way to leverage connectors that others have built, like the Kafka connect type stuff. >> Truthfully we are a heavy-duty user of open source software so our actual product, if you dig in to what you see, it's a framework for executing pipelines. And it's for connecting other software into our product. So it's not like when we integrate Kafka we built a build brand new blue sky Kafka connector. We actually integrate what stuff is out there. So our idea is to bring as much of that stuff in there as we can. And really be part of the community. You know, our product is also open source. So we play well with the community. We have had people contribute connectors. People who say we love the product, we need it to connect to this other database. And then they do it for us. So it's been a pretty exciting situation. >> We were talking earlier off-camera, George and I have been talking all week about the badge workloads, interactive workloads, now you've got this sort of new emerging workloads, continuous screening workloads, which is in the name. What are you seeing there? And what kind of use cases is that enabling? >> So we're focused on mostly the continuous delivery workload. We also deliver the batch stuff. We're finding is people are moving farther and farther away from batch in general. Because batch was not the goal it was a means to the end. People wanted to get their data into their environment, so they could do their analysis. They want to run their daily reports, things like that. But ask any data scientist, they would rather the data show up immediately. So we're definitely seeing a lot of customers who want to do things like moving data live from a log file into Hadoop they can read immediately, in the order of minutes. We're trying to do our best to enable those kind of use cases. In particular we're seeing a lot of interest in the Spark arena, obviously that's kind of why we're here today. You know people want to add their event processing, or their aggregation, and analysis, like Spark, especially like Spark SQL. And they want that to be almost happening at the time of ingest. Not once it landed, but like when it's happening. So we're starting to build integration. We have kind of our foot in the door there, with our Spark processor. Which allows you to put a Spark workflow right in the middle of your data pipeline. Or as many of them as you want in fact. And we all sort of manage the lifecycle of that. And do all those connections as required to make your pipeline pretend to have a Spark processor in the middle. We really think that with that kind of workload, you can do your ingest, but you can also capture your real-time analytics along the way. And that doesn't replace batch reporting for say that'll happen after the fact. Our your daily reports or what have you. But it makes it that much easier for your data scientists to have, you know, a piece of intelligence that they had in flight. You know? >> I love talking to someone who's a practitioner now sort of working for a company that's selling technology. What do you see, from both perspectives, as Spark being good at? You know, what's the best fit? And what's it not good at? >> Well I think that Spark is following the arc of like Hadoop basically. It started out as infrastructure for engineers, for building really big scary things. But it's becoming more and more a productivity tool for analysts, data scientist, machine-learning experts. And we see that popping up all the time. And it's really exciting frankly, to think about these streaming analytics that can happen. These scoring machine-learning models. Really bringing a lot more power into the hands of these people who are not engineers. People who are much more focused on the semantic value of the data. And not the garbage in garbage out value of the data. >> You were talking before about it's really hard, data movement and the data's not always right. Data quality continues to be a challenge. >> Bryan: Yeah. >> Maybe comment on that. State the data quality and how the industry is dealing with that problem. >> It is hard, it is hard. I think that the traditional approach to data quality is to try and specify a quality up front. We take the opposite approach. We basically say that it's impossible to know that your data will be correct at all times. So we have what we call schema drift tools. So we try to go, we say like intent-driven approach. We're interacting with your data. Rather then a schema driven approach. So of course your data has an implicit schema as it's passing through the pipeline. Rather than saying, let's transform com three, we want you to use the name. We want you to be aware of what it is you're trying to actually change and affect. And the rest just kind of flows along with it. There's no magic bullet for every kind of data-quality issue or schema change that could possibly come into your pipeline. We try to do the best to make it easy for you to do effectively the best practice. The easiest thing that will survive the future, build robust data pipelines. This is one of the biggest challenges I think with like home-grown solutions. Is that it's really easy to build something that works. It's not easy to build something that works all the time. It's very easy to not imagine the edge cases. 'Cause it might take you a year until you've actually encountered you know, the first big problem. The real, the gotcha that you didn't consider when you were building your own thing. And those of us at StreamSets who have been in the industry and on the user side, we've had some of these experiences. So we're trying to export that knowledge in the product. >> Dave: Who do you guys sell to? >> Everybody. (laughing) We see a lot of success today with, we call it Hadoop replatforming. Which is people who are moving from their huge variety of data sources environment into like a Hadoop data-like kind of environment. Also Cloud, people are moving into the Cloud. The need a way for their data to get from wherever it is to where they want it to be. And certainly people could script these things manually. They could build their own tools for this. But it's just so much more productive to do it quickly in a UI. >> Is it an architect who's buying your product? Is it a developer? >> It's a variety. So I think our product resonates greatly with a developer. But also people who are higher up in the chain. People who are trying to design their whole topology. I think the thing I love to talk about is everyone, when they start on a data project, they sit down and they draw this beautiful diagram with boxes and arrows that says here's where the data's going to go. But a month later, it works, kind of, but it's never that thing. >> Dave: Yeah because the data is just everywhere. >> Exactly. And the reality is that what you have to do to make it work correctly within SLA guidelines and things like that is so not what you imagined. But then you can almost never go backwards. You can never say based on what I have, give me the box scenarios, because it's a systems analysis effort that no one has the time to engage in. But since StreamSets is actually instruments, every step of the pipeline, and we have a view into how all your pipelines actually fit together. We can give you that. We can just generate it. So we actually have a product. We've been talking about the StreamSet data collector which is the core like data movement product. We have like our enterprise edition, which is called the Dataflow Performance Manager, or DPM, It basically gives you a lot of collaboration and enterprise grade authentication. And access control, and the commander control features. So it aggregates your metrics across all your data collectors. It helps you visualize your topology. So people like your director of analytics, or your CIO, who want to know is everything okay? We have a dashboard for them now. And that's really powerful. It's a beautiful UI. And it's really a platform for us to build visualizations with more intelligence. That looks across your whole infrastructure. >> Dave: That's good. >> Yeah. And then the thing is this is strangely kind of unprecedented. Because, you know, again, the engineer who wants to build this himself would say, I could just deploy Graphite. And all of a sudden I've got graphs it's fine right. But they're missing the details. What about the systems that aren't under your control? What about the failure cases? All these things, these are the things we tackle. 'Cause it's our business we can afford to invest massively and make this a really first-class data engineering environment. >> Would it be fair to say that Kafka sort of as it exists today is just data movement built on a log, but that it doesn't do the analytics. And it doesn't really yet, maybe it's just beginning to do some of the monitoring you know, with a dashboard, or that's a statement of direction. Would it be fair to say that you can layer on top of that? Or you can substitute on top of it with all the analytics? And then when you want the really fancy analytic soup, you know, call out to Spark. >> Sure, I would say that for one thing we definitely want to stay out of the analytics base. We think there's many great analytics tools out there like Spark. We also are not a storage tool. In fact, we're kind of like, we're queue-like but we view ourselves more like, if there's a pipe and a pump, we're the pump. And Kafka is the pipe. I think that from like a monitoring perspective, we monitor Kafka indirectly. 'Cause if we know what's coming out, and we know what's going in later, we can give you the stats. And that's actually what's important. This is actually one of the challenges of having sort of a home-grown or disconnected solution, is that stitching together so you understand the end to end is extremely difficult. 'Cause if you have a relational database, and a Kafka, and a Hadoop, and a Spark job, sure you can monitor all those things. They all have their own UIs. But if you can't understand what the is on the whole system you're left like with four windows open trying to figure out where things connect. And it's just too difficult. >> So just on a sort of a positioning point of view for someone who's trying to make sense out of all the choices they have, to what extent would you call yourself a management framework for someone who's building these pipelines, whether from Scratch, or buying components. And to what extent is it, I guess, when you talk about a pump, that would be almost like the run time part of it. >> Bryan: Yeah, yeah. >> So you know there's a control plane and then there's a data plane. >> Bryan: Sure. >> What's the mix? >> Yeah well we do both for sure. I mean I would say that the data point for us is StreamSet's data collector. We move data, we physically move the data. We have our own internal pipeline execution engine. So it doesn't presuppose any other existing technologies, not dependent on Hadoop or Spark or Kafka or anything. You know to some degree data collector is also the control plane for small deployments. Because it does give you start to stop commanding control. Some metrics monitoring, things like that. Now, what people need to expand beyond the realm of single data collector, when they have enterprises with more than one business unit, or data center, or security zone, things like that. You don't just deploy one data collector, you deploy a bunch, dozens or hundreds. And in that case, that's where dataflow performance manager again comes in, as that control plane. Now dataflow performance manager has no data in it. It does not pass your actual business data. But it does again aggregate all of your metrics from all your data collectors and gives you a unified view across your whole enterprise. >> And one more follow-up along those lines. When you have a multi-vendor stack, or a multi-vendor pipeline. >> Bryan: Yeah. >> What gives you the meta view? >> Well we're at the ins and outs. We see the interfaces. So in theory if someone were to consume data out of Kafka do something right. Then there's another job later, like a Spark job. >> George: Yeah. >> So we don't automatic visibility for that. But our plan in the future is to expand as dataflow performance manager to take third party metric sources effectively. To broaden the view of your entire enterprise. >> You've got a bunch of stuff on your website here which is kind of interesting. Talking about some of the things we talked about. You know taming data drift is one of your papers. The silent killer of data integrity. And some other good resources. So just in sort of closing, how do we learn more? What would you suggest? >> Sure, yeah please visit the website. The product is open source and free to download. Data collector is free to download. I would encourage people to try it out. It's really easy to take for a spin. And if you love it you should check out our community. We have a very active Slack channel and Google group, which you can find from the website as well. And there's also a blog full of tutorials. >> Yeah well you're solving gnarly problems that a lot of companies just don't want to deal with. That's good thanks for doing the dirty work, we appreciate it. >> Yeah my pleasure. >> Alright Bryan thanks for coming on "The Cube." >> Thanks for having me. >> Good to see you. You're welcome. Keep right there buddy we'll be back with our next guest. This is "The Cube" we're live from Boston Spark Summit. Spark Summit East #SparkSummit right back. >> Narrator: Since the dawn.

>> Announcer: theCUBE presents On the Ground. (ambient techno music) >> Hello, everyone. Welcome to a special theCUBE presentation of On the Ground here at Oracle's corporate headquarters. I'm John Furrier, the host of theCUBE, I'm here with Chai Pydimukkala, senior director of product management with Oracle. Welcome to On the Ground, appreciate you coming on. >> Thank you very much. >> So, talk about the data integration strategy and plans for Oracle, and what are some of the products that make that up? >> Oracle data integration, we've been around for more than 15 years. We've been helping our customers to move data from various systems, sources, and targets. Our products consist of a real-time data integration product, which is used for continuous availability of real-time replication, which is Oracle Golden Gate. It's our marquee product, it's been around for two decades. We also have a ETL product called Oracle Data Integrator, which is a product that actually takes the data, and then, it transforms the data in the source and the target itself. It's not like the older technologies, where you pull the data out of the system and process it in a middle tier. Instead of that, we actually leverage the power of the source of the target. And that's where we started. We have a data quality suite and a complete data governance foundation. We have about 12,000 customers, you know, talk about the largest banks, largest telcos in the world. Each and every one of them use our product, so that completes our data integration product portfolio. >> So, what is this new data integration cloud suite we've been hearing about because that's interesting, ties into that? Does that relate and how does that play? >> Absolutely, so what we have done is one of the things that we have been focused as Oracle is, we have had so much traction in the cloud space, so we have seen that when customers are moving their database systems or applications or platforms into the cloud, one of the key challenges remains is how do you get that data from on-premise to cloud, or cloud to on-premise. That's where data integration comes into play, and what we have done is we have taken the existing technologies that we have, like our Golden Gate, like Oracle Data Integrator, and data governance foundation, and we are making it as a part of a solution stack that gets available, that gets provisioned in cloud, so that any customer can come in and get these products, Oracle cloud integration stack, data integration stack, and then, they can start doing moving data from on-premise to cloud, or cloud to on-premise, or pure cloud use cases. And the stack that we are envisioning is we are not only looking at our traditional products that we have, like Golden Gate, which is a replication product, and ODI, Oracle Data Integrator, but we are also introducing couple of new products. One is Dataflow machine learning, which I'll talk about it in detail, and then, we also have a data-wrangling product called Big Data Preparation Cloud Service, which is already launched and available today, where people are going to look at data and start doing semantic extraction of the data. That's the biggest announcement is our customers will be able to come to us, and instead of focused on the real-time use case or a batch use case, they'll be able to get a solution stack, a platform, that they can use for data integration, be it real-time or be it batch or be it application integration or database integration. >> What's this Oracle Dataflow ML, machine learning thing about, Chai, because that's also kind of a new thing that's coming up? >> You know, I think one of the things that we have done at Oracle is we have been in the forefront of innovation, so a lot times we do solve enterprise level machine critical use cases, but one of the things internally that we have done is we have been embracing, constantly embracing, real-time and open source technologies, big data technologies, and cloud technologies. One thing that we observed in the marketplace is the traditional ETL is like driving a car using your rear-view mirror. You're not actually analyzing the data as it's coming in, you're actually have moved the data, transformed the data, and looking at the data, and started making decisions. Instead of doing that, what we think is we have built a new platform where we can analyze data as it's flowing through. So, let's say your transactions are coming in. You want to detect any fraud on your transactions, banking transactions, what we can do is now we can feed the data, capture the data using Golden Gate and feed it into this engine called Dataflow Machine Learning engine, and then, we'll be able to do a lot of fraud analytics in real-time on it. The whole paradigm of the batch ETL versus real-time ETL is evolving right now, and what we are introducing is a platform that's completely built on an OpenStack Spark-based platform. We are leveraging natural language processing and machine learning, so that as the data comes in, be it your transactional data, be it any other seeming data, we can actually look at the data and give you more insights in real-time so that either you can create alerts or events, or you can detect fraud, or you can actually get more insights and do transformation on the data and make it available to your business. >> How much does open source play into this? You mentioned that. A lot of people always ask me that, so I had to ask you. >> One of the things that we have consistently have managed to do is not to reinvent the same thing again and again. For example, when we actually talked about, envisioned about Dataflow machine learning, the technology itself, we had one thing in mind that we did not want to introduce another engine. If you look at the traditional ETL companies that are going obsolete right now, they're introducing their own engine where they feed the data into this engine. But what we think is the future is that this open source community is so rich, and there are so many people are working on it, we need to leverage those contributions. For example, our Oracle Data Integrator never had an engine, so we followed the same principle, and even in Dataflow, we don't have an engine, we use the Spark libraries, we use the machine learning capability, we use the algorithms from natural language processing, excuse me, and then, we actually combine all this information and we can process them natively on a Hadoop platform, which is the open source platform. And then, lo and behold, you can get more insights into your-- >> You're not restricting customers. You let them do whatever they want with the data if it's connected in, say, a big data appliance, and, or cloud suite. >> Yes. >> So, you kind of give them the choice. >> Yes, so, one thing that we have done very consciously at Oracle is, we acknowledge Oracle database as the number one database in the world. We have more than 50% of the enterprise customers, Fortune 500 customers, actually almost all of the Fortune 500 customers use us, right? But the point is we also realize that there are all these other heterogeneous sources where people have been using to store data. The polyglot architecture where people store graphs in a graph database or NoSQL key value pairs in a NoSQL type of database is valid, and we understand the use cases. So, all the product capabilities-- >> They're not mutually exclusive. A database now can be put where the data makes sense. >> Exactly. >> But you guys just still be the systems of record. >> Yes. >> 'Cause you're the CRM, the ERP, you have all these data systems that are powering business. >> Absolutely, so. >> Why would you restrict data coming in, right? >> Exactly, so one of the things that companies want to do and customers want to do is they want to be able to take the mission-critical transaction data that they have, and they want to be able to combine it with the social media data or the interaction data that they're getting, or the weblogs data, and they want to be able to correlate the information and get more insights. If you look at it like, you know, if you look at customer experience, if you want to really know your customers, what they are doing, you want to get the CRM data, which is their mission-critical data, but you also want to combine it with the social networking data, what do they like, what are they interacting with, what are they clicking on the website, so that you can combine both. We have been a heterogeneous platform, we have customers, we have got a customer who actually uses us only for non-Oracle systems, which is absolutely fine with us. We are in the business of data integration. We do it very well with Oracle technologies, but we can also support other technologies. >> I mean, you guys don't ask customers to be Oracle database everywhere, but in the key areas you do. The question I have to ask you is the one I get all the time from customers and people out in the field, practitioners, and I'm going to paraphrase kind of the pattern question. Oracle, you guys are amazing on the database side, but I want to just integrate other data sources, and I don't want to have to buy Oracle. That's what I'm looking for. What are you doing, Oracle, to make your database smarter? Because their, the customer's view is, okay, I've got Oracle database, you know. Can I get out of that swim lane and expand the intelligence of the Oracle database to a Hadoop, to a Spark, to another environment? >> We have done a lot of-- >> How do you address that? >> We have done a lot of innovation in terms of database, I just think data management in general. First of all, on the data integration side, we have had customers, the largest cell phone company in the world, moves data from an Oracle database to a Kafka-based queue to do further analysis. The largest electric car manufacturing company is actually trying to optimize their assembly lines in real-time so that they don't lose money if their assembly line goes down. We have done a lot of innovation where, and a lot of these customers are using big data type of technologies to get additional insight, so we don't stop them from taking data out from Oracle database or putting data back into Oracle database. Not only that, what we have introduced is. >> You're encouraging people to move data fast around to and from Oracle. Why not, right? >> Exactly, because if you want to get more insights, you want to combine all kinds of data, your interaction data, your NoSQL data, your weblog data. We are saying that bring it in, you can use a big data platform. We have an offering called Big Data Appliance cloud, Big Data Appliance, and we are offering it as a cloud service, too, where you can actually take an Oracle database, and you can take a big data system, and we can connect it, and we have connected it with NoSQL, with Big SQL adapters, so that you can issue SQL, and it can operate on both these sets of data. >> Operationally, that's a really easy way for a customer, rather than deploying a separate system, training assist admin. >> Exactly. >> Cost of ownership is probably going through the roof. >> Absolutely >> Do you see that as a key enabler? >> Absolutely, absolutely, and I think we are in the business of data integration. We treat all data sources and targets equally, and we'll try and support because when people are, when customers are making this journey to the cloud, it's important that we treat everybody equally. >> The old joke that we have, Dave Vellante and I on theCUBE, we say if customers wake up from a coma from 10 years ago and they're in today's world, and the data warehouse is all different, what do you say to that person? Well, welcome back to the real world, but I mean, that's the kind of awakening that these enterprises are having, where a lot of people haven't made the investment, but now are under a lot of pressure to modernize. They know Oracle database, they've had some great relationships, but now all of a sudden the world has changed. What do you say to those folks, what is the most compelling thing that's changed over the past five to 10 years, that's happening now that didn't happen then? >> I think the two big pivots that we have had in the industry are the big data pivot, where people are looking at multiple data management systems and the big data pivot, and then, the cloud pivot because cloud is very important, and we have seen our customers, we have been helping our customers to move entire data center into the cloud, in Oracle public cloud infrastructure, where they are saying I don't want, I want to reduce my total cost of ownership, improve productivity, I want to get all these tools that are already available out there, and I don't want to install this software on my system. Data warehouse as an analytical store will still exist, but what's happening is the transition where you move this data, transform the data, where you transform the data, and where you create operational data stores is changing, and that's where we come in and we say, if you have a big data system, you can create your operational data store over there, transform all the data over there and send it to your warehousing system. We are not, you know, we, because data warehousing again it's post-analysis. It's not real-time analysis as the data is flowing in, so I think, and then the cloud, you know, all we have made sure that for our customers, all the platforms that are available today, we have both infrastructure as a service platforms, SaaS-based service, and we also have data as a service, we are making sure that all these innovation platforms that we have created, including data integration, are available to our cloud customers. Anybody who wants to go to the cloud, and they want to get away from these other, older mainframe systems, they can come in and use our data integration technology, use our database, use our big data appliance cloud service, and just pivot to the cloud immediately, and don't have to wait. >> So, speed to the cloud, speed to a modern architecture. If I hear you correctly, you're saying that Oracle's philosophy and strategy is to have the best modern data management system given the customer's best choice. >> Absolutely. >> Would that be a fair statement? >> Absolutely. And to add to that-- >> Of course, buying some Oracle database, but using open source if they want to. >> Absolutely. >> Where the tool makes sense. >> Because one of the things that we have done on our cloud is we not only offer our platforms, we also offer big data platforms. If you want Kafka as a service, it's going to be available. Spark as a service, it's available. We have embraced Docker. A lot of these things are available. >> How 'about the competition, where do they stand compared to Oracle? >> You know what, can I say, I spent 10 years at a competitor, and then, I made the change, I joined Oracle three years ago, and that competitor is not even a public company anymore. On the data integration space, we have dominated, we have grown. We have got about 12,00 customers and it's growing. We are adding new logos everyday. >> John: And what's the difference, why is that, why are you guys competitive? >> Because the three things that we are focused on is no engine, so we did not invest in an engine for our transformation, so we don't pull in the data and transform it in our engine, that's one. Second is real-time. We are focused on real-time because we know that the future is people will want to analyze this data in real-time, so our real-time platform, which is Golden Gate platform, is world-class and it's the number one platform. And the last one is we make this, everything, we make it easily available in the cloud and for big data platforms. So, you don't have to change anything, it's fairly simple. >> Chai, thanks for spending some time with me on the ground here at your headquarters. >> Thank you very much. >> I'm John Furrier here, exclusive coverage of Oracle here On the Ground with theCUBE. I'm John Furrier, thanks for watching. (light electronic music)