 From the CUBE studios in Palo Alto in Boston, connecting with thought leaders all around the world, this is a CUBE Conversation. Hi everybody, this is Dave Vellante, and welcome to this CUBE Conversation. Really excited to have Wim Kokertz in. He is the senior vice president of software development at Oracle. Wim, it's great to have you on. And you know what? I often say I wish we were face to face, but if we were, you'd have to cut off my tie because developers and ties just don't go together. No, I know, and this is my normal app. So this is me wherever I go. So hi again, good to see you. Yeah, great to see you. So of course, a lot of people are confused about Oracle and open source, they say Oracle open source. What is that all about? But I think you misunderstood. People don't, first of all, realize you as the leader of the software development community inside of Oracle. I mean, you've been involved in Linux since the early 90s, but you guys have a lot of committers. You do a lot. I want to talk about that. What is up with Oracle and open source? Yeah, well, it's a broad question. So, you know, a couple of things. One is we have many different areas within the company that are dealing with open source, right? So we have the cloud team doing a lot of stuff around cloud SDKs and support for different languages like Python and Go and of course Java and so forth. So they do a lot around ensuring that the Oracle ecosystem is integrated in the open source tools that customers use or developers use, terraforms and so forth. And then you have the Java team and so of course Java is open source. And then the Grail project, Grail VM, which is a polyglot compiler that run Java and Python and JavaScript and so forth together in one VM do really cool optimizations. That's an open source project, also on GitHub. There's of course MySQL, which is, you know, along with Java, they're probably the two most popular and widely used open source projects out there. There's VirtualBox, which is of course, also a very popular project that's open source. There's all the work we do around Linux. And I think one of the things is that when you have so many different areas doing things that are for that area, then as a developer or as a customer, you typically just deal with that group. And what you see is, oh, you're talking to the Java developers. So you know what's going on around Java. The Java developers might not necessarily say, oh, and we also do MySQL and we do Linux and VirtualBox and so forth. And so you get sort of a rather myopic, narrow view of the larger company. When you add all these things up and there would be one big slide that says, this is Oracle. These are all these open source projects there. And there's multiple ways, right? One is we have projects that we've open sourced and all the code came from us and we made it publicly available. We are the main contributor and we get contributions back. There are other projects where we contribute to third party in terms of enhancing things, like I said, with the cloud team. And then in general, something like Linux where we're part of an external project and we participate in the development of that project at large. And so there's these three different ways. When you count up all the developers that we have that deal with open source on a daily basis and in terms of contributions in terms of bug fixes, testing and so forth, it's thousand literally full time developers. And of course, all the projects are on either on GitHub or similar sites that are very popular. So yeah, I think the misunderstood is probably a lack of knowledge of the grasp of what we do. And our primary goal is to provide services and products to customers. And so the open source part is sort of embedded in the development methodology, but it's not a, that's not something we sell or market separately. We just, we work with customers and products and services. And so in some cases, it's not well understood. Yeah. Well, we're talking, of course, we're talking about the state of the penguin. I think it's important for people to understand that Oracle got into the Linux game in the 90s, maybe the latter part of the 90s. And Oracle, of course, wants to make Linux, wants to make Oracle its applications and database run better on Linux. But as you're pointing out, your Linux distro, full support end to end, thousands of people in your open source community. And the contributions that you make to Linux, many, if not most, they go upstream. Everybody can benefit from those. But of course, you want an Oracle distro that is going to make Oracle stuff run better. That's always kind of been the Oracle way. Well, so, yes, two things. The one is that, so everything we do is upstream. So we have no Linux patches that are not contributed upstream. There's no proprietary code in Oracle Linux at all. It's all completely open, publicly available. The source code, the change log, all the commits, everything is fully open and public, which sometimes is not well understood, but it's completely open. And everything we do in terms of feature development or functionality or bug fixes goes upstream to the Linux kernel mail list. It's actually, it's the only way to be able to manage a Linux distribution and be a Linux vendor is to live in that ecosystem. Otherwise, the cost of maintaining your own fork, so to speak, is very high and it doesn't really solve the problem. Now, the functionality we work on obviously is focused on making Oracle products run better, making Oracle Cloud run better and so forth. However, what again, what's important to understand though is in Oracle Database is a program running on an operating system, it does IO, it does networking, it does memory, it deals with memory management, lots of processes. So for the most part, the things we work on to improve that helps everyone else, right? It helps every other database run better or helps every other language run better. So none of these changes are specific to Oracle. There's just things that we found doing performance benchmarks and testing and so forth. But we say, hey, if Linux did the following, it would make boot up fast. Now, boot up has nothing to do with the database. But if our customers run on one terabyte, four terabyte, eight terabyte systems, and so booting up and Linux starting up and cleaning up memory takes a long time. So we want to reduce that from an availability point of view. So here, we're now talking about just enterprise, right? And so there's this broad set of things we work on that definitely help us, but they're actually really completely generic and help everyone else. Yeah, that's great, good. So I wanted to kind of get that out of the way and help our audience understand it. So let's get into it a little bit. What are you seeing? What's going on in IT? What, pick your observation space and your vision of what you see happening out there. Well, you know, it's very interesting. It's sort of, there's two worlds, right? There's the cloud world and the move to cloud and there's the on-premises world where people run their systems on their own. And one of the things that we've learned is, you know, when you talk about machine learning, obviously is something that's very popular these days and automation. And so in order to rely on machine learning well and have algorithms that are very effective, you need lots of data. And so being a cloud vendor and having Linux in our cloud on tens of thousands or hundreds of thousands of servers or more allows us to have a view of how an operating system works across incredibly large scale. So we get lots of data. And so for us to figure out which algorithms work well in terms of how can we do network optimizations, how can we discover anomalies on the storage side and deal with it and so forth, we can do that at scale. And what's interesting is how do we then bring that to on-prem? Well, if we can get the data and the learning done, the training done in our cloud directly, then when we provide that service also to people running Oracle Linux on-premises, then that will work. The alternative is to have point solutions where you provide something to a customer and it needs to learn something from small amount of data. That doesn't work so well. So I think having both worlds on-prem and cloud directly allows us to kind of benefit from that. And I think that's important because lots of customers are interested in going to cloud. Many of the enterprises have not yet, you know, they're starting but there's still a huge on-premises space that's important. And so by being able to get familiar with how these things work at scale, autonomy is again important, right? Autonomous database is incredibly popular and so forth. That allows us to then say here, try these things out here. It's a service. We can show you the benefits right away. And then as that improves, we bring that on to a certain extent on-premises as well. And then they can have it in both places. And that I think is something again, that's relatively unique but also very important is that we want to create and we want to provide services and product that acts similarly on-premises as well as in cloud because at some point when people move, we want to make that transition seamless. And what you have today for the most part is one world that's on-prem and then the cloud world is completely different. And that is a big barrier of moving. And so we want to reduce that. You can run the same operating system local as well as cloud, you can get the same functionality. And then that helps transition people over much. Yeah, well Oracle actually was one of the, I think about Oracle was the first company to actually market same-same. You actually use that term. Others put forth that concept but Oracle was the first to announce products like cloud, a customer that was same-same. Now it took some time to actually get it perfective and get it to market. But the point is, and we've written about this, is that Oracle, because of the ascendancy of cloud, flipped and has a cloud-first mentality. And you just kind of referenced that. You just said, and you can bring that to on-prem. So I wonder if you could talk about that cloud-first mentality and the impact on hybrid. So yeah, I think the cloud-first part is, of course, in cloud we work with services more so than products that we deliver. And there's a number of things that are happening. So one is we obviously continue to provide products that we can download Oracle Linux, you can download the database and weblog you can install it on your app, right? You can do the traditional way of working. Then in a cloud world, what typically happens is, oh, I use a database service. I'm not installing anything. I push a button and I get an IP address and a SQL neck and neck string and I connect to the database. And we take care of everything underneath with autonomous database. Now, in order to do that, you need a whole infrastructure, right? You need logging agents. You need a backend that captures all that stuff. You need monitoring tools. You need all the automation scripts for bringing this service up and monitor it. And so that takes a lot of time to do, right? And we learn a lot by doing this. And so the cloud-first part of these services means that we get to experience this ourselves with direct access to everything. Now, taking that service with all of the additional features like autonomy and bringing that to an on-premises world, we have to make sure we can package that so that all these pieces around and go along with it. And that takes a little bit more time. So we can't do everything at the same. And so what we've done with autonomous databases, we traded everything in Oracle Cloud. We have the whole system running really well. And then we've been able to package that and shrink it into something that can be installed on-premises, but then connect it into Oracle Cloud again. And so that way we can get all the telemetry, all the metrics. And that allows us to scale because part of providing a cloud service that runs on-prem in the customer environment is that we need to be able to remotely manage that similar to how we manage something that runs in our own cloud, right? Otherwise it doesn't scale. And so that takes a little bit of time, but we've done all that work and now with cloud-accounted database that's really in place. Yeah, you really want to have that same cloud experience, whether it's on-prem, you know, in the public cloud, hybrid, et cetera. So I want to explore a little bit more who is using Oracle Linux and what's the driver for using it. Can you describe maybe some of the types of customers and why they buy? Sure, so we started, you know, 14 years ago, right? 2006, October 25th, 2006. Remember that day very well. Penguin's on stage and a big lunch for Oracle Linux in San Francisco at the Moscone Center. So look, the initial driver for Oracle Linux was to ensure that Oracle database customers or Oracle product customers had a good operating system experience, right? And the ability to be able to handle critical issues when that occurs because typically a database runs the company's critical data. The most essential stuff that a company has is typically in a database, an Oracle database. And so when that thing has issues with the operating system, what you don't want is to talk to multiple vendors and have finger pointing and having to explain to an operating system vendor how the database works. In the Unix world, we had a good relationship with the OS vendors and the hardware vendors. They were the same and they knew our products really well. And in the Linux world, that was very different. The OS vendor basically did not want to understand or learn anything about the products living on top. And so while to a certain extent that makes sense, it's in an enterprise world where time is of the essence and downtime needs to be limited, absolutely. We can't have these arguments and such. And so that was the driver initially for doing Oracle Linux. It was to ensure there was a Linux distribution fully backed by us that we could fix and we could fully support. That was completely the original intent. And so the early customer base was database customers, database and middleware, mostly database. So, but that has then evolved quickly. And so, sorry, what happened was people run, people would say, look, I have a thousand servers, a hundred run Oracle. So we'll run Oracle Linux on those hundred and we run something else on those other 900. Now, after a year or so, they realized that our support was really good. We fix all these issues. And so then they're like, why are we having two links distributions? This thing works really well. It runs any application, it's fully compatible. So we'll just go a thousand with Oracle Linux. And so the early days, the first few years was definitely Oracle Database as the core driver. And then it sort of expanded to the rest of the, to the rest of the estate. And over the years, you know, we've added lots of features and functionality like case supplies and so forth. We have an attractive pricing model for running on servers. And so now lots of our customers have a very small Oracle percentage running and many other things running. So it's really become a all or nothing play in the Linux space. And we're well known now. So it's been actually very good. You just mentioned case splice. We've been talking about cloud and on-prem and hybrid. And let's talk about security because security really is a differentiator. But particularly if you're going to start to put stuff into the cloud, talk about case splice specifically but generally security and your posture there. So security first is sort of, you know what you hear us say and do in everything we do, right? The database obviously security on the Linux side, security matters. Case splice as a technology is there to do critical bug fixing and make sure that we can apply security vulnerability fixes without affecting the customer and not have downtime. Right? And if you look at, you know, most of the cases or many of the cases where you have security vulnerabilities and exploits, it tends to be because systems were not patched. Why were they not patched? Well, not that a customer doesn't understand that it's important, but it's a whole train of events that needs to happen. You have to, you know, you get notified that this is security issue and you're operating some more application. Then, well, an application typically means it's a multi-tiered setup. So if you have to bring your database server down then you first have to coordinate with the application users to bring the app server down because that talks to the database. So to patch one system, you basically have to bring down whole application stack. You have to negotiate with the DBAs. You have to negotiate with the app admins. You have to negotiate with the user. It takes weeks to do that and find time. Well, during that time, you're vulnerable. So the only way really to address security in a scalable way and reducing that window of time is to do it without affecting the customer. Right? And so Case Splice is something that, it's a company we acquired in 2009 and have since evolved in terms of capabilities. And so it allows us to patch the Linux kernel without downtime, right? We locked the kernel for eight microseconds. It's literally no downtime. You don't have to bring down applications. The user doesn't see it. There's no hang, there's no delay. And so by doing that, you can run the Linux operating system or Linux and you can be fully patched on a system that hasn't rebooted for three years. And you don't even know it. And so by doing that type of stuff, it makes customers more secure and it avoids them. It saves them a lot of money in terms of dealing with project management and so forth, but it really keeps them secure. And so we do that for the Linux kernel. We do that for some of the libraries on top that are critical like OpenSSL and GLFC. And, you know, one example, I can give you two examples. So one example is Heartbleed was this bug in OpenSSL a number of years ago. And so everyone had to patch their SSH server. And that meant, you know, basically systems around the world had to reboot, like a whole IT reboot across the world. With K-Splies today, if Heartbleed were to happen tomorrow, we would be able to patch this online for all the Oracle Linux customers without any downtime. No reboots, no restarting of applications, everything that's running. The amount of money saved would be massive, right? And also of course the headache. Another example is, and this was in Oracle Cloud, when some of these CPU bugs that happened a few years ago that were rather damaging on the cloud side, right? Where you could basically see memory of potentially of other machines running, but the cloud is incredibly critical. We were basically able to patch our entire cloud within four hours, and the customer didn't know, right? 120 million patches or something that we applied within four hours, all online without any downtime. And so that technology has been really helpful, both for us to run our cloud, but the exact same patches and same fixes go to customers on premises as well. But this comes back to the whole, what we do in cloud, we also do for customer. And I think that's a unique thing that we have at Oracle, which is quite fascinating, right? The operating system we run for our customers, the operating system that's the host for the VMs is the exact same binary and source code that we make available, just to be clear. The exact same binaries are the ones that you run as a customer on premises. So if you run Oracle Linux with KVM, you run VMs, you're actually running the exact same stuff as we do for our, that we run underneath our customer stuff. Nobody else does that. Everyone else has a black box. So I think that helps a little bit with transparency as well. Yeah, and that homogeneity just creates an environment and you're talking about sort of that security mindset. It's just critical, you're not just bolting it on. It's part of the culture. But you were started your career, and then of course you were a Linux person when you came to Oracle, but then I think you spent some time in database back in the day when there were some serious database wars going on before Oracle became the king of database. So now you've got obviously this great portfolio and a lot of really sharp software developers. What should we expect going forward from Oracle? What should we look for? You know, I was talking to some, I was welcoming some interns to the company for their summer internship yesterday. And one of the things that I, I'm sorry, one of the things I mentioned to them was that one of the, so cloud obviously gives us a lot of opportunities, but there's a number of things. One is we have such a breadth of applications and software and hardware together, right? We have the servers, we have the storage, we have the operating systems, we have the database layer and so forth, and we have the cloud side. And one of the great opportunities, and I think we've shown a lot of this happening with the ability to create something like autonomous database is to combine all these things, right? We have such a broad portfolio of really cool technology that by itself is okay, but if you combine the things, it really becomes awesome, right? You cannot create autonomous database without having autonomous linux, right? You cannot create those two and make them really safe without also controlling the firmware on the hardware and so forth. So by being able to combine all these layers and by having a really great relationship across the teams within the company, that opens up a lot of opportunity to do stuff really quickly and having the scale for that. I think that has been for the last few years a really great thing, but I can see that being one of the advantages that we have going forward, right? We have Oracle Fusion applications, which is incredibly popular and has great growth and then we have that running on Oracle Cloud, it talks to our autonomous database. So we bring all these pieces together and no other SaaS vendor can do that because they don't have these other pieces. They have one area, we have all of them. And so that's the exciting part for me is basic. It's not so much about making my own world better and having linux be better and case flies and so forth, which is important, but that becoming part of the bigger picture. And that's the exciting part. Well, Oracle's always invested in R&D. We've made that point many, many times, whether it's database, Fusion was a painful, but worthy effort, the whole public cloud piece, many obviously many acquisitions, but the investments that you've made in open source as well, Wim, you're a great spokesperson and a great representative of the open source community and generally in an Oracle specifically. So thanks very much for coming on theCUBE and sharing with us the state of the penguin and best of luck. You're welcome. Thank you, thanks for having me. All right, and thank you for watching everybody. This is Dave Vellante for theCUBE. We'll see you next time.