>>Welcome welcome back to the cubes coverage of dr khan 2021. I'm john for a host of the cube. We're here to talk about observe ability in the enterprise, enabling developers. Fabian lang VP of engineering and co founder of Istana, now part of IBM. Fabian, Congratulations on everything and great to have you on the cube here for dr gone. >>Thank you. Thanks for having me. >>So I'm in Palo Alto, you're in Germany were doing the remote thing obviously virtual second year in a row for dr khan. Soon real life is coming back. Uh no real impact of developers as they continue to be more productive than ever. The hottest conversation topic being discussed, being funded by venture capitalists and private equity is observe ability. This is an area you guys are playing in aggressively and you got some product observe ability. What's the big deal about Docker con Docker containers observe ability kubernetes, Why is observe ability at the center of all these conversations and the center of the value. >>So observe ability basically means you understand what's going on and today it's more important than ever to understand what's going on because there is so much more going on. If you think back five years maybe before Dr even was featured prominently, you had very little things that you needed to control that you need to understand and then micro service and coordinative became more popular and became really more important to understand what all those moving parts are doing. And that's where observe ability was born out of what we have been doing before at that time it was called application performance monitoring A PM. It's now called observe ability. It's really understanding all those parts of your architecture, of your stack of the application and in the end of the end user experience, you want to know if a user is experiencing a slow service and what's the reason for that? Because today, so many things are moving so many things that maybe even outsourced into cloud providers, it's more important than ever to know what's going on. >>Well we're here at Dunkirk on 2021 virtual. I want to get you to take a minute if you don't mind explaining to the folks why Dr and Dr Khan is important to Astana. >>So I, I said we were founded like six years ago and at that time Doctor was the rising star. It was promoting a lot of new technology. It was giving developers new abilities to develop applications in a very agile away. Microservices were enabled by Doctor before you had to deploy those things somehow it was a city Rome and then you needed to install >>debian >>package but with microservices you have so many more things to install. So it was really, I would say instrumental to the success of microservices to have a platform like docker that was really the next gen of technology that helped to enable those applications. And for us it was really an important driver to understand the whole stack, the traditional tools where eyes are oriented to infrastructure monitoring. So you understand the quality of your host if it's running slow or to look into application of an application was throwing errors but everything was disconnected and unique functionality of Astana is to connect all those bits and pieces of the application together and for that containers. And now kubernetes is a really important part to understand because it is part of this whole picture. >>Did you talk about the problem that you guys solve? Um obviously with those availability, I mean the general concept, we kind of get that great, great overview on your part, but when you start to get into devoPS teams, you start looking at def sec off, start looking at cloud native applications. I see Docker containers provides all that goodness and kubernetes, orchestration, etcetera. What problem do you guys solve? And um what's the benefit? >>The main problem that in stana solves is getting all this understanding that I said is required to provide a good experience of to your users, to your end customers uh without requiring you to do all the instrumentation work or the capture and configuration work because in stana is very automatic, it automatically sees all the works lords that are running in your communities, for example, that are running in Dr containers, but it also connects to legacy databases, fully automatic. So no configuration required also means that with a high rate of change that some of those applications hard have is that we will see all those change happening in real time. And you can't forget to make a configuration to enable your observe ability. So it's really return of investment on the viability solution that we provide and we provide a lot of this insight uh that you can get and that enables you to provide better service for your users. >>So you guys aren't just a doctor monitoring service and company, you guys actually run on Docker. Right, is that true? >>That's correct. So we are not only monitoring doctor and all these things connected to applications, but we are running on a doctor or platform as a service. SaAS software as a service. We run for you so you don't need to operate and stana, we are running it on managed kubernetes clusters and uh, IBM cloud and amazon cloud in google cloud. We have all that and it's it's all running on docker containers and that gives us so many features that are really great with DACA. So all the configuration that specific to microservices are being baked into the images and you can just roll it out, especially for monitoring products that is dependent on the data, that the performance depends on the data our customers send. Um, these ease of scalability with doctor is just so much bigger than it would be with a traditional deployment type. We can just add worker notes to our cluster and have ports auto scale to new notes and this is functionality that wasn't there before and that's great and that's important, essential for our business. >>You know, one of the conversations that's being talked about here at dr khan and in the industry at large is this idea of happy developers and everyone wants to keep developers happy. I've been hearing that conversation, have many chats with folks, you know, productivity and innovation, um but productivity and happy developers of the concept, but also, you know, on the, on the business side or on the developer side, it's more accelerated pipeline. Right? So, so how do you manage to flow, keep that productivity going, But also enabling happy developers, what do you guys do to help there? I mean what if someone asks you, hey, how do you make my developers happier and accelerate my pipeline? >>Well, that's really dependent on what makes the developers happy. I think most developers really want to get their functionality. They are working on their passionate about into production into the hands of end users. So um, skipping out a lot of the manual configuration work that's boring and not really appealing to develop us, helps everything is pre packaged and configured automatically. So that's a big, big plus. And the standard monitoring as I said, uh, is also automatic. So you don't need to configure it, your, your application on how to monitor it. So developers can just focus on delivering features and whenever there is something we will tell them, I think they enjoy that >>innovations creates great, that's a benefit. Can you talk about the on prem version of installing a, that's something that you guys are talking about and featuring um what is that about? Can you take a minute to explain beyond prem version of in Astana for dr containers? >>Yeah, it's a, it's an interesting topic, especially at the conference like dr khan, where it's all about virtualization, container realization and going into the cloud, that there are still companies, enterprises government mental entity that are very heavily invested on an on premise solution. They want to have control or are legally required to have control over what they have been deployed. So we knew when we founded in Astana that our solution, unlike our competitors, can't be only software as a service. We want to have a fantastic software as a service product and experience, but it should be equally good on premises as well. And when we were looking at ways how to actually do it, how to deliver an architecture that a little bit complicated to on premises customers to have themselves as the solution. We saw that doctor solves a lot of problems for us. We don't need to manually petra around operating system that customers, we don't have different versions of packages installed. It's all the same and actually it's not only all the same for all the deployment of all our customers, but it's also the same technology that we run as a software. As a service customers can run it now on their own. So we have feature parity, it's not lagging behind and this is also ease of support for us. >>So why was it, what was the motivation behind that was just customer demand? Um, more efficiency? What was the motivation behind moving on, supporting the on prem version? >>Uh, so for a start up, it's all about addressing the market share. Right? So you wanna have everything you can get, you don't want to spend any extra money on it. And as I said, the enterprise market is big. There are still many players that want to have the data in house. This is potentially sensitive data that's being tracked. So an on premise solution having, it was really instrumental to the success of in Stana because we were able to target and help those customers even in a fully adapt scenario, for example where they don't even have internet access. >>Take me through the process of DACA rising the product sitting on prime product that you get the thing going on there, like okay, let's do this. What does that look like? How did that work out? >>So as I said, we looked at this from the beginning and we picked DACA as a technology from the beginning, so there wasn't really like a shift and left type of scenario that other customers might be having. We were doing it from the beginning and we were aligning our architecture so that there are no fundamental differences between an on premise solution and anti size solution. That's of course configuration, that's different. But that configuration we just put into a single configuration file and that turned out to be a great idea because this is how you nowadays configure your application kubernetes, you'll make a customer resource for example, and then have an operator run the product, any kind of product, but also in stana, you run on premises with an operator that just works on the single configuration that you give it. And this is actually great because our customers are used to operating products like that, their own software, everything customers are running in dhaka in kubernetes, they are used to operating it that way. And that helped us because our customers now get the same functionality that we offer as a, as a service on premises very easily very quickly. And that make them happier. We talked about developer happiness that makes them happy because now they are not lagging behind but it also enables us to give better quality support, lot fixes faster and helps us to no longer support very old presence because they don't exist. They are frequently updated. I think this is really a benefit of container realization is also how easy it is to upgrade because you just stop apart and start a part in the new version and then you have a new verse. >>That's also great insights may be great to chat with you on that. I got to ask you on a personal note, you've been in the industry for a while and your leader, um you know, that's a performance geek, you'll have to build fast code. I was been chatting with other VPs of engineering and we were talking about the shift in engineering and with devops you've got kind of s our reaction, you have some just straight up application coding, just modernize that cloud native applications and you've got a kind of under the devoPS as the world's shifts. It seems like there's more of an architectural systems engineering approach or a systems mindset and that seems to be changing the mindset of a developer from Iterate fast. And then the line I heard was you can iterate and pump out code fast, but it might not be good, might be crap. So, so this notion of iterating code and crafting good product because with now this module Ization with containers, you're doing a lot more design work. So craft seems to be coming back to coding. Uh, I don't think it's coming back, it's been there, but it just seems more of like, hey, let's do this, right? And it's not just ship code. What's your take on that? >>So I think this always was there. It's just that traditionally companies approached software engineering similar to how companies approach manufacturing. So somebody writing a designs back and somebody verifying it and then it's going onto the line to mass production. But software doesn't work that way. We make way more changes, it's way harder to understand it up front. So the developed the iterative and exile development that has been ongoing is really, is really what people want and develops well. There is this notion of being a being waking up in the middle of the night and that's what developers don't want. So you need to prepare your application, you need to make it resilient against that. And developers are very eager to build in functionality that helps them to troubleshoot to make their application available. With a high rate of change. There is a high rate of risk as you said and I think the ability to deploy 1000 times per day is great but you don't necessarily need to do that. I think it's also important for your users that you find the right pace of when you deliver functionality and when you deliver fixes. >>I was just talking to a friend the other day and we were just talking about organizations and teams and yeah, we always riff on the the two pizza team or having more agility and you have this democratization because of the agility is also a benefit for any developer to add value if they have the right perspective or creativity. But it kind of disrupts the kind of the old way of thinking. I'm the principal engineer is my job. No, I'm the chief architect. So you have these titles and you have roles, the roles are changing and sometimes just the arguments. Oh wait, that's my job is that I'm this kind of changes. What's your thoughts on, how do you manage that dynamic? Because as you have more, uh, I won't say surface here more democratized engineering with virtual teams and whatnot You have compose ability with, with with code. You have more of a systems are a lot more going on. It's not your standard engineering mindset. What's your thinking on this as a leader in engineering and visionary? >>Well as we know the architecture of a software full of the organization that the company has. That creates. All right. So I think what you want when you want to have a micro service architecture, you want to have a micro service teams. You want to have teams, we call him at and standard delivery teams that work more or less independently on a certain set of features and are responsible for them and to end. So my engineers, they are talking to our customers figuring out how to make a feature better. They are then designing this with our user designers and then they are developing and deploying it and this really entry and responsibility. And we don't really have those titles like architect anymore. I think those roles are still there but it's more like a shared responsibility. So you of course want an architecture, you want to have your components talk to each other in an efficient way and it's more really communities of practice that are establishing. So you will find out that you have people and your teams who have specific skills who like to work on architecture. Some of them like to work on continuous delivery systems And then you you form those cross functional teams dynamically and when it's no longer hit this bands. And I think that's a major difference to assigning a person to a road. >>Yeah and and also that with you have new trends like observe ability, enterprise observe ability you know new things are happening um And new net new things like new architecture and also new roles and responsibilities. I'll see new patterns to with the data you have services being stood up and turned down all the time. You have a lot of dynamic environment. So you know having a happy developers one eliminate the manual work what you do but also giving them good work assignments to work on some good hard problems. So what is what are those hard problems that engineers like to work on these days? Is it like design? Is it coding? I mean I know it depends as you mentioned on the personalities but generally speaking as dev ops def sec Ops becomes much more of an agile edge hybrid play. What's the hard problem? >>I think big data is not really a new term but I think this is still a very interesting territory because you can apply various aspects to it. You have this data science aspect to it to understand how to detect pattern in it. And then automation is actually artificial intelligence. Right? So you automate data science and that's very interesting because those are large scale problems and new problems and new solutions. So yes there are existing frameworks but there's so much innovation to be found and making this work efficiently is another dimension of the same problem. That's also not easy and challenging problems. Make developers happy and then you can even have people think about the financial aspects. So it should also be cheap Big data and AI is usually very expensive because it requires so much hardware. So not only tried to make it fast but maybe even make it efficient. So this whole domain is very appealing. There is new technology to be invented, tough problems and I think that's really exciting to developed. >>Fabian Lang, vice president of engineering co founder and stand a great to have you on the q Great insight. Thank you for sharing that knowledge there. And the overview of installing here at dr khan observe ability very relevant for next gen next level solutions. Thanks for coming on the cube. Right, okay. I'm john Fury with the queue here. Dr khan 2021 coverage. Thanks for watching. Mm.

(string music) >> Welcome to the Cube conversation here at the Cube studios in Palo Alto California, I'm John Furrier, cohost of the Cube and cofounder of Slip and Angle Media. We're here with Shlomi Ben Haim who's the founder and CEO of Jfrog, hot startup. I asked him to come in to chat about his business. In the dev/op space, we see him at a lot of shows, your company's doing well, we love the marketing, the frog thing is great, love it, very cool. But there's a lot of real serious action going on in the enterprise and in the cloud and in emerging tech, whether it's AI or machine learning, whether its innovative things, developers are front and center in the marketplace and there's a boatload of noise out there, there's like this approach, this approach, there's a lot of different approaches, but at the end of the day, the devs are driving a lot of innovation. You guys are at the center of it so welcome to the Cube. >> Thank you. >> First question for you, just take a minute to talk about what you guys do, Jfrog, what's your company, what's your business, what are you guys up to, what's your deal? >> The way I think that the community will describe us would be that we are the binaries people, we are taking care of your binaries. As you know in the dev/ops world, everything you do you do with your binaries, with your software artifacts, so I heard some of the community members call us the database of dev/ops and we are the providers of artifactory, bintray, xray and mission control which take care of your binaries, managing them, host them, distribute them and secure them. >> Open source event we were at, we saw you guys because I was doing all the interviews and you guys were right on the edge there, then you guys got some nice background images off the Cube videos, but it was really interesting. The trend is your friend as the saying goes and the number of open source projects is increasing, the actual lines of code is exponentially going to grow from 22 million to 200, 400 million lines of code over the next couple of years, that's hockey stick. More developers are coming in, not old school like me that built their own stuff from scratch, there's a lot of lego blocks in fact Jim Samlin said that 10% of the code will probably be original ideas and differentiation, 90% of most of the code will be a code sandwich, which I believe, I think that's the legit direction. How do you guys fit into that trend and what does that mean for your business because I can imagine, there's a ton of Git Hub stuff going on, tons of forking, tons of projects, you got block chain catching the world by storm, there is a massive developer tsunami going on. How do you guys help them? >> It's very interesting, when we started Jfrog, actually my co-founder Yoav Landman started by providing developers with a very dummy, basic solution to proxy, public repositories like Maven central and it was not about the code for the first time, it was about the binaries. Code is great and the line of code, as you said, it's going to go enormous but what happened is that when you need to automate, when you need to rebuild, when you need to release faster, you go down to the binary level, to the software artifact level and guess what, no one took care of your binaries before, you were just throwing your binaries to your version controller or file store, maybe you were hosting them. >> They were messy, it's like a kid with their room, all the stuff spread around all over the place, where's that binary, no one keeps track of it. >> Nobody care about that, but this is the one thing that you keep consuming, keep building with, keep recompiling and in the era of dev/ops, is the one asset that you need to automate and reuse. This is where we, >> The core problem if I get this right, is that compiling is going to be, if you think of dev/ops, it's infrastructure as code, as the phrase goes as we always say and programming infrastructure is what dev guys want to do, they don't want to be in the business of switching configurations, getting in the routers and the network. They want it to be just one layer of resource, serverless is a great trend for you, more and more developers are going to love this. They want to program, so when you're programming, the inherent next step is where's the code, who's compiling it, does it need to be compiled? Is that the core problem, that there's more and more stuff going on under the hood that needs to be managed? Is that growing part of your business solution or is the problem just lost binaries, what's the core problem? >> It's a perfect question. First of all, we are providers, we are the providers of the only universal solution. Binaries are not just for java developers, they are not just for python developers, they are not just for dot net developers, they are not just for docker users and the way you package it, binaries happens between your get and your CI server, let's say Jenkins, get and Jenkins and your Kubernetes. Something happens between those two sites. Your orchestration tool and your code repository tool. In this land is where binaries play a very significant role and this is where we are a major player. >> Is the problem error prone in that zone, in that zone it's like the wild west, it's like a black hole if you will, think about what you're saying, if I get it right. There's a lot of stuff that goes on in there, is it mismanagement, what's the core thing that you guys got to do there? >> Tons of binaries, too much public repositories that the community cannot rely on. You need to manage and host your own binaries, the ones that you create yourself, and to provide and this is the last strength we see in the market, big organization need to provide dev/ops as a service to their own developers, so they need to ask this very important asset that we call software artifact and binaries or darker images or whatever you want to call it. >> Yeah a lot of great trends going on, obviously containers and Kubernetes you mentioned. Let's get into those, that's driving a lot of change. Certainly containers has been around for a while, whether you call it wrappers or whatever, it's a great magical thing, we love containers, Kubernetes really gets the trend right, if you look at the google trend, you see Kubernetes has got so much more traction than containers, although I'm not saying one's more relevant than the other, certainly orchestration's important, linking and loading all these containers together. Why is Kubernetes accelerating the binary conversation? Is it because more rapid development is going on, more programmability's going on, why is Kubernetes impacting the binaries components more now than ever? >> Putting aside the need for automating and integrating, this whole orchestration solution requires some work on the binary level but if you think about what Kubernetes is trying to solve, or what the containers are all trying to solve is a better, faster release, better, faster deployment, better, faster delivery and then you can do it only if you will combine the power of the developers and the power of the machine and release faster. This is what we say in Jfrog, release fast or die because it's all about how fast can you release? >> Before we get into some of the product specific stuff, I want to ask you some pointed questions on that. I want to ask you about automation. AI is obviously hot, I love AI, even though it's hyped up, it still promotes great software development, machine learning really is where the meat on the bone is there, so machine learning and automation bots, whatever you want to look at it, is an opportunity to actually to create adaptive code. How did that new software paradigm affect binaries because I can almost imagine that if you got a bot going wild, it could screw up the binaries. >> Completely. >> So can you comment on that, that area. Obviously we want more bots, automation is a good thing on one level, but how do you guys look at that market as an opportunity, as a challenge, what's that whole AI thing look like? >> Well if we take a step back, I think the dev/ops started with the need to automate and release faster. It was like the playground of developers, we need a better integration, we need a continuous integration, we need better delivery, we need continuous delivery. If you think about it now, in 20/20 perspective, you understand that this was all milestones. The next big challenge is continuous updates. People like me, people like you, just want their devices and machines to be updated. >> And secure, look at Equifax. Equifax is a great example, they didn't update the code. >> Absolutely and it's flowing and just happening and secure and in the world of automation, the world of AI, I think that the big challenge or the next big challenge of dev/ops is how can I create a continuous update machine which is also secure and software update will just flow. It will not be something that you press I agree, I reboot and do any kind of crazy stuff in order just to get your software update. It's more about the user experience of all of us. It's not just developers and dev/ops companies anymore. >> That's a great vision by the way, I love that. It should work like that and programmable infrastructure for dev/ops should be programmable and always available and highly reliable. Mark Zuckerberg used to have the saying, move fast, break stuff, that's not a dev/ops ethos by the way, they built their own dev/ops, but then he kind of quickly waffled back to move fast, be reliable, because he got some religion on ops. Totally get that, let's go into today's world. That gives us a little future view, what is a use case for a customer? Take me through the day and the life of a customer that's using Jfrog, what are their problems, what are some of the things that are burning in their office? Where's the smoke, what's the problem that they have that they need to take care of the binaries? Sprawl of code, just mismanagement, what are some of the signals? Share with your view there. >> It starts with the fact that it's not your developer anymore that builds software. You have a CI server there that never goes to a lunch break, never take a break with Facebook, which by the way, it's a great company but sometimes it stop giving the time during the work time and you keep building and building like crazy. Your CI server keep producing binaries. >> It's an always on code machine basically. >> It's a binaries machine and it's being built 24 by seven and yes, you use just a portion of it but you have to host and manage all of it and if you will host it in your version controller, it will explode, if you will put it in a file store, it will not be something, >> Explode because of capacity? >> Because you cannot do any cleanups on the version controller, not get or subversion or the false or any of them, you don't do cleanups on version control. >> Hygiene is an issue. >> Yes, plus integration. You need to integrate with your records system, plus promotion, you need to allow and automate promotion of the specific bites that you build. >> So that's why people call you the database or I would even say the brains of binaries, you got to keep track of the goods if you will, it's like the crown jewel is the binary. >> Right. >> If I get that right, okay let's take it to the next level. You have good hygiene, you have good stuff going on, what are you guys doing specifically that gives you a differentiation of the market because is it software, is it hardware, what is the Jfrog differentiation? >> I think that the first thing that happened to us was that we realized that binaries is for everyone. If you remember Jfrog's slogans from 2010, it was binaries for the people. We felt like we are leading the revolution of taking care of your binaries and therefore, we decided that whatever we build, our philosophy base, our concept will be universal. We started with the Java community, Maven and Gradel and then the dot net community with Nougat and then when it came to be more like a dev/ops industry in 2013 or '14 was it, >> Roughly, 2008 to 2014 was really the cloud errati and then it grew and then it matured a little bit. >> And the combination of dev and ops and IT and then we started to support packages like Debian and RPM, beyond repositories, docker registry, we were the first docker registry in the market. >> You were riding the wave from the beginning. >> Yes. >> You were right there riding the binary wave with the native cloud growth, public cloud growth big time which by the way had a lot of iterations quickly. >> Which is also one of our differentiators, we are the only hybrid providers for your binary solution. We have it in the cloud, any cloud or on prem. >> Who's the competition? >> It's a very good question, on a niche level, we have companies like docker that provide a docker registry we have Cores that provide docker registry, by the way, anyone in the market now that want to have a docker registry, a container registry. On the Java Maven domains, Sonotype provide a nexus which is a binary repository manager for Java for Maven builds. NPM provide a solution for NPM but if you think about the universal solution that supports other, >> Those are siloed platform specific binaries. >> Yes. >> You're taking much more of a wholistic, horizontally scalable, any binary any time management. >> Exactly, we don't do the before and after, but in the binaries world, we want to be one solution for all. >> I get the whole registry thing, love that positioning. Just a dumb question, when someone's coming in and managing intermittently in the registry, how do you guys handle that piece? How do you know that a Java request coming in from a Java registry, you guys have a front end to this thing, is it your software, how do you guys manage the integration of requests to and from the binaries. >> The read and write to the repository you mean? >> Yes. >> Artifactor is a very sophisticated repository if I may say it's built more like a database, it's based on a check sum mechanism and not just a basic file store. >> You verify it coming in on the front end. >> Right, the parts and machine caching, managing, hosting and distributing, it's all happening in artifactor. >> And performance is as good? No problems with performance? >> Well we are the only provider that has a highly available solution with over 4000 customers, so I guess it is. >> You got a smile yeah, I see you at the shows. You got a good reputation so it's great to have you come in. I want to just take a minute to pause because I know we're having a great conversation, I could talk about CI servers til the cows come home, one of my favorite topics dev/ops, as people who have been following me since 2008 know, I love the cloud, cloud native vision from day one. There's a lot of people out there who don't know what the hell a binary is, so take a minute and explain, what is a binary and why is it such an important thing right now in context to open source growth, more developers coming in, context to enterprises trying to be cloud like and just for the general purpose, why are binaries important? Why should the general public, how would you talk about what is a binary? >> I'll try. I think that the main difference is that binaries are more like, maybe it's a basic metaphor, but binaries are more like fresh food, unlike freeze food. Your source code is freezed, you're not allowed to touch it, you're not allowed to clean it, you're not allowed to change it. Your one dot seal would be my one dot seal. It's kind of freeze food, this is why in dev for get and other player in this market are so important. We see how bit bucket with the class in and Git Hub are growing and still playing a significant role binaries are different, binaries is the fresh food. Something that you keep changing, any minute and you build with a specific binary something and then something else and it become another binary if I may say so. I think that the flexibility that you need to gain when you go on full automation and full integration is the flexibility that you can get on the binary level. You cannot get it on the code level. Therefore, binaries playing a very significant role in the cloud era and in the dev/ops era. >> Sure it allows for extensibility of source code. In a way what you're saying. You can eat the frozen food or you can chop up your own organic meal yourself. >> Exactly. >> Okay I get that, final question for you, thanks for coming in, appreciate the one on one on binaries there. People can always just go on Wikipedia and look at other definitions on stack overflow and whatnot. What is the customer value proposition for Jfrog? Why should I work with you, what's the main reason for you to have 4000 customers? What's driving them to use you? Is it just convenience? Is it scalability, all of the above? Just take a minute to explain why customers go to you and if people don't work with you, why should they work with you? >> I think that the biggest challenge today is that you want to treat binaries as first level citizens and instead of having an NPM repository, docker registry, Maven repository, python repository and there is no single organization that will have just one repository, you can have it all with Jfrog. The second thing we are the providers of highly available solution to protect your data centers so if you don't want your 1000 developers sitting down, waiting for the binary repository to be up and running and to allow the environment, then you probably want to, >> Productivity right there is one. >> Productivity and efficiency, absolutely. We are also providing it to secure your binary flow and the platform that distributes your binaries. We take binaries very seriously, over two billion downloads a month on bintray, our distribution hub and we work with the community and for the community, we are developers ourself, coming from the open source community so it's all bottom up and community friendly. >> Shlomi, great commentary, I want to just get a personal, take your Jfrog hat off for a minute, put your developer, executive, industry expert hat on. Share with the audience your view on the developer market. There's been a lot of negative press around the brogrammer lately and all these things, but trends are clear that you have massive growth in open source, comment on the role open source plays as it goes into some argue fifth generation, fourth, fifth generation, I remember when the first generation I was coding on. Those were the days but different, it's changed. You have so much code, it's really a party right now in open source, there's so much good stuff happening. Google's donating tensorflow, all these people putting real big libraries out there to code on. Kubernetes is just so awesome, system guys specifically love what's going on in the cloud. But cloud is exploding a lot of opportunities, IoT and AI, what's the developer market like right now, just share your thoughts, what's the sentiment, what's the excitement, what are the young kids doing? What are some of the big things that you see happening? >> From business perspective, what we see in the market is developers first of all taking decisions. They hear their managers coming with the pain and expect it to solve it and the bottom up process is something we never saw in the market. The last five, six years, we see more and more developers kind of educating their managers with how to do it and how to do it faster. The second thing and this is, >> So bottom up's happening now you're saying. >> Happening for the last five years and it's growing. The second thing we see in the cloud, you see it more than I am, Google and Amazon and Microsoft and Red Hat, everyone want a piece of the cloud, Orca now just announced two days ago, three days ago. Everyone want a piece of the cloud and everybody understand that data traffic comes from developers, it's not individuals, it's communities, the open source community is giant and it's a very, there's a very important player in the data traffic of what we call the cloud highway. >> And the communities are very most important piece, you would agree with that, right? We're very community focused, that's the key, right? >> Yes, absolutely. >> I think the world will be developer indoctrinated with basically developer premises across all business, so it's not a department anymore, it's permeating all through organizations. >> Right and also impact our user experience. People like simple people that doesn't understand code, they're not contributing to the open source world still need software updates and competitive analysis are talking about that, how fast can you release? >> Well Stu Miniman and Dave Alante and Peter Burris and I always talk about community is the key in open source, you guys have been very successful in the community. Congratulations, obviously we're very community focused with our content, with the Cube. If you like the Cube, check us out at cube.net, give us a call, come in the studio if you're a thought leader, love to chat with you. I'm John Furrier with the Cube, more thought leadership coverage in Palo Alto here inside the Cube. We'll be right back, thanks for watching. (electronic music)

>> Announcer: Live from Manhattan, it's theCube, covering AWS Summit, New York City, 2017. Brought to you by Amazon Web Services. >> Welcome back to the Big Apple as we continue our coverage here on theCube of AWS Summit 2017. We're at the Javits Center. We're in midtown. A lot of hustle and bustle outsie and inside there, good buzz on the show floor with about 5,000 strong attending and some 20,000 registrants also for today's show. Along with Stu Miniman, I'm John Walls, and glad to have you here on theCube. And Dustin Kirkland now joins us. He's at Ubuntu, the product and strategy side of things at Canonical, and Dustin, good to see you back on theCube. >> Thank you very much. >> You just threw a big number out at us when we were talking off camera. I'll let you take it from there, but it shows you about the presence, you might say, of Ubuntu and AWS, what that nexus is right now. >> Ubuntu easily leads as the operating system in Amazon. About 70%, seven zero, 70% of all instances running in Amazon right now are running Ubuntu. And that's actually, despite the fact that Amazon have their own Amazon Linux and there are other, Windows, Rails, SUSE, Debian, Fedora, other alternatives. Ubuntu still represents seven out of 10 workloads in Amazon running right now. >> John: Huge number. >> So, Dustin, maybe give us a little insight as to what kind of workloads you're seeing. How much of this was people that, Ubuntu has a great footprint everywhere and therefore it kind of moved there. And how much of it is new and interesting things, IOT and machine learning and everything like that, where you also have support. >> When you're talking about that many instances, that's quite a bit of boat, right? So if you look at just EC2 and the two types of workloads, there are the long-running workloads. The workloads that are up for many months, years in some cases. I met a number of customers here this week that are running older versions of Ubuntu like 12.04 which are actually end of life, but as a customer of Canonical we continue providing security updates. So we have a product called Extended Security Maintenance. There's over a million instances of Ubuntu 12.04 which are already end of life but Canonical can continue providing security updates, critical security updates. That's great for the long-running workloads. The other thing that we do for long-running workloads are kernel live patches. So we're able to actually fix vulnerabilities in the Linux kernel without rebooting, using entirely upstream and open source technology to do that. So for those workloads that stay up for months or years, the combination of Extended Security Maintenance, covering it for a very long time, and the kernel live patch, ensuring that you're able to patch those vulnerabilities without rebooting those systems, it's great for hosting providers and some enterprise workloads. Now on the flip side, you also see a lot of workloads that are spikey, right. Workloads that come and go in bursts. Maybe they run at night or in the morning or just whenever an event happens. We see a lot of Ubuntu running there. It's really, a lot of that is focused on data and machine learning, artificial intelligence workloads, that run in that sort of bursty manner. >> Okay, so it was interesting, when I hear you talk about some things that have been running for a bunch of years, and on the other side of the spectrum is serverless and the new machine learning stuff where it tends to be there, what's Canonical doing there? What kind of exciting, any of the news, Macey, Glue, some of these other ones that came out, how much do those fit into the conversations you're having? >> Sure, they all really fit. When we talk about what we're doing to tune Ubuntu for those machine learning workloads, it really starts with the kernel. So we actually have an AWS-optimized Linux kernel. So we've taken the Ubuntu Linux kernel and we've tuned it, working with the Amazon kernel engineers, to ensure that we've carved out everything in that kernel that's not relevant inside of an Amazon data center and taken it out. And in doing so, we've actually made the kernel 15% smaller, which actually reduces the security footprint and the storage footprint of that kernel. And that means smaller downloads, smaller updates, and we've made it boot 30% faster. We've done that by adding support, turning on, configuring on some parameters that enable virtualization or divert IO drivers or specifically the Amazon drivers to work really well. We've also removed things like floppy disk drives and Bluetooth drivers, which you'll never find in a virtual machine in Amazon. And when you take all of those things in aggregate and you remove them from the kernel, you end up with a much smaller, better, more efficient package. So that's a great starting point. The other piece is we've ensured that the latest and greatest graphics adapters, the GPUs, GPGPUs from Invidia, that the experienced on Ubuntu out of the box just works. It works really well, and well at scale. You'll find almost all machine learning workloads are drastically improved inside of GPGPU instances. And for the dollar, you're able to compute sometimes hundreds or thousands of times more efficiently than a fewer CPU type workload. >> You're talking about machine learning, but on the artificial intelligence side of life, a lot of conversation about that at the keynotes this morning. A lot of good services, whatever, again, your activity in that and where that's going, do you think, over the next 12, 16 months? >> Yes, so artificial intelligence is a really nice place where we see a lot of Ubuntu, mainly because the nature of how AI is infiltrating our lives. It has these two sides. One side is at the edge, and those are really fundamentally connected devices. And for every one of those billions of devices out there, there are necessarily connections to an instance in the cloud somewhere. So if we take just one example, right, an autonomous vehicle. That vehicle is connected to the internet. Sometimes well, when you're at home, parked in the garage or parked at Whole Foods, right? But sometimes it's not. You're in the middle of the desert out in West Texas. That autonomous vehicle needs to have a lot of intelligence local to that vehicle. It gets downloaded opportunistically. And what gets downloaded are the results of that machine learning, the results of that artificial intelligence process. So we heard in the keynotes quite a bit about data modeling, right? Data modeling means putting a whole bunch of data into Amazon, which Amazon has made it really easy to do with things like Snowball and so forth. Once the data is there, then the big GPGPU instances crunch that data and the result is actually a very tight, tightly compressed bit of insight that then gets fed to devices. So an autonomous vehicle that every single night gets a little bit better by tweaking its algorithms, when to brake, when to change lanes, when to make a left turn safely or a right turn safely, those are constantly being updated by all the data that we're feeding that. Now why I said that's important from an Ubuntu perspective is that we find Ubuntu in both of those locations. So we open this by saying that Ubuntu is the leading operating system inside of Amazon, representing 70% of those instances. Ubuntu is, across the board, right now in 100% of the autonomous vehicles that are running today. So Uber's autonomous vehicle, the Tesla vehicles, the Google vehicles, a number of others from other manufacturers are all running Ubuntu on the CPU. There's usually three CPUs in a smart car. The CPU that's running the autonomous driving engine is, across the board, running Ubuntu today. The fact that it's the same OS makes it, makes life quite nice for the developers. The developers who are writing that software that's crunching the numbers in the cloud and making the critical real-time decisions in the vehicle. >> You talk about autonomous vehicles, I mean, it's about a car in general, thousands of data points coming in, in continual real time. >> Dustin: Right. >> So it's just not autonomous -- >> Dustin: Right. >> operations, right? So are you working in that way, diagnostics, navigation, all those areas? >> Yes, so we catch as headlines are a lot of the hobbyist projects, the fun stuff coming out of universities or startup space. Drones and robots and vacuum cleaners, right? And there's a lot of Ubuntu running there, anything from Raspberry Pis to smart appliances at home. But it's actually, I think, really where those artificially intelligent systems are going to change our lives, is in the industrial space. It's not the drone that some kids are flying around in the park, it's the drone that's surveying crops, that's coming to understand what areas of a field need more fertilizer or less water, right. And that's happening in an artificially intelligent way as smarter and smarter algorithms make its way onto those drones. It's less about the running Pandora and Spotify having to choose the right music for you when you're sitting in your car, and a lot more about every taxicab in the city taking data and analytics and understanding what's going on around them. It's a great way to detect traffic patterns, potentially threats of danger or something like that. That's far more industrial and less intresting than the fun stuff, you know, the fireworks that are shot off by a drone. >> Not nearly as sexy, right? It's not as much fun. >> But that's where the business is, you know. >> That's right. >> One of the things people have been looking at is how Amazon's really maturing their discussion of hyrid cloud. Now, you said that data centers, public cloud, edge devices, lots of mobile, we talked about IOT and everything, what do you see from customers, what do you think we're going to see from Amazon going forward to build these hybrid architectures and how does that fit in to autonomous vehicles and the like? >> So in the keynote we saw a couple of organizations who were spotlighted as all-in on Amazon, and that's great. And actually almost all of those logos that are all-in on Amazon are all-in on Amazon on Ubuntu and that's great. That's a very small number of logos compared to the number of organizations out there that are actually hybrid. Hybrid is certainly a ramp to being all-in but for quite a bit of the industry, that's the journey and the destination, too, in fact. That there's always going to be some amount compute that happens local and some amount of compute that happens in the cloud. Ubuntu helps provide an important portability layer. Knowing something runs well on Ubuntu locally, it's going to run well on Ubuntu in Amazon, or vise versa. The fact that it runs well in Amazon, it will also run well on Ubuntu locally. Now we have a support -- >> Yeah, I was just curious, you talked about some of the optimization you made for AWS. >> Dustin: Right. >> Is that now finding its way into other environments or do we have a little bit of a fork? >> We do, it does find it's way back into other environments so, you know, the Amazon hypervisors are usually Xen-based, although there are some interesting other things coming from Amazon there. Typically what we find on-prem is usually more KVM or Vmware based. Now, most of what goes into that virtual kernel that we build for Amazon actually applies to the virtual kernel that we built for Ubuntu that runs in Xen and Vmware and KVM. There's some subtle differences. Some, a few things that we've done very specifically for Amazon, but for the most part it's perfectly compatible all the way back to the virtual machines that you would run on-prem. >> Well, Dustin, always a pleasure, >> Yeah. >> to have you hear on theCube. >> Thanks, John. >> You're welcome back any time. >> All right. >> We appreciate the time and wish you the best of luck here the rest of the day, too. >> Great. >> Good deal. >> Thank you. >> Glad to be with us. Dustin Kirkland from Canonical joining us here on theCube. Back with more from AWS Summit 2017 here in New York City right after this.

>> Robert: It feels like we're talking because it's boring TV. Tech people love tech. Consumers love the benefit of tech. No consumer opens up their iphone and says oh my gosh, I love the technology behind my iphone. >> What's it been like, being on the Shark Tank? >> You know filming is fun and hanging out is fun and it's fun to be a celebrity at first. Your head gets really big and you get really good tables at restaurants. >> Who says tech isn't got a little pizazz. >> Voiceover: More skin in the game. In charge of his destiny, Robert Herjavec is Cube Alumni. Live from Santa Clara, California, it's the Cube covering Open Networking Summit 2017. Brought to you by the Linux foundation. >> Hey, welcome back everybody, Jeff Rick here with the Cube. We're at the open networking summit 2017 in Santa Clara, California. I think it's the fourth year of the conference. We've been coming for a long time. It's pretty amazing, a lot of transformation is happening as this project moves from the conversational to the testing to actual. A lot of deployments being talked about in the keynotes. So happy to have Scott Raynovich joining me again. >> Pleasure as always, thank you. >> Did you have a good night last night? >> Excellent. >> Alright, super. Super guest Drew Schulke, he is the vice president of converged networking at Dell EMC. Drew, Welcome. >> Thanks, thanks for having us. >> You've been busy at this show. You're doing panels, you're doing keynotes, they're working you. >> It's been a bit of a whirlwind going on thus far. Yeah, I had a keynote talk on economic and organizational impacts of open networking, which went really, really well. A lot of great questions from the audience, really insightful questions on that. Have met with folks like yourself, some other folks in the media, some analysts, talking to some customers which is always nice to have. We'll close it out tomorrow with one of the keynotes. A panel discussion on the role of open source and moving to 5G where I'll be participating with some folks from Intel and Ericsson. I'm looking forward to that, but yeah, definitely a whirlwind week. >> So before we get into some of the specifics, just your general impressions as to how this thing has evolved over time. Impressions of this show this year. >> Yeah, great question. I think the thing that struck me the most this year was the amount of customers coming in and actually talking about putting a lot of the things we've been talking about at this summit for several years into production environments and seeing results out of it. Some great keynotes last night by some folks, Amadeus in the travel industry and talking about their journey actually moving things into production, I thought was fabulous. Which gives people a vision of what really is possible and moving these out of the theoretical and here's the vision, the strategy into here's how you can actually get things done and getting into results. Ultimately, when you put things into production, that's how you ultimately learn and refine things over time. It's a great move forward for us. >> Awesome, so on the economics and organizational impact of open networking, your keynote. What are some of those really key economic drivers that you outlined in that conversation? >> Yeah, great question. You can kind of break it into a capex and an opex discussion. On the capex side, what we've seen is this whole open networking model is built on merchant silicon and the commoditization of hardware, which may sound like okay, that's a bad thing. Well no, it's really, really good because what it's doing is it's allowing all of us to take the benefits of huge volume and scale that's going on. From the biggest cloud providers down to the enterprise, as we move into this hardware model that's based upon merchant silicon and more standard network designs that are capable of supporting multiple OS's, we all benefit from the economies of scale that go in that. We can amortize R&D investments over a larger number of things. That's all goodness, so there's a huge tailwind on the capex side. On the opex side, as we start to disaggregate the network stack and focus on the individual layers, it creates a different operational model that allows for a high degree of automation. One of the things that we brought up in the session was contrasting a study from 2013 where the typical enterprise network admin was controlling about 300 ports. That was the breadth of support that they had back in 2013. That same year, Facebook came out and said an individual operator can support up to 20,000 servers. It's not like they're just super humans. What happened in there was a level of automation. That's a key ingredient of our open networking strategy, is driving that automation. That's where you get true economies of scale on the opex side. Those are the main points on that one. >> Jeff: Yeah, good ones. >> So Drew, one of the themes we've seen here is that the Linux foundation has done a good job of consolidating some of the open source technology and putting them in the same place so we can all track them and figure out what's going to happen. You just told us about Dell donating some of your code to the Linux foundation. >> Drew: Correct. >> Why don't you explain how you made that decision and what you think it's going to do for your customers. >> Yeah absolutely, as a little bit of context, what we see happening in terms of networking software is one, it's become decoupled from the hardware. That's already done right now. But even when we start to look at the software side, we think there's more disaggregation possible. We can peel apart the layers of what currently is a network operating system today and create a based operating system upon which several different companies can come in and put in what at that point becomes applications on top of it to do things like an L2, L3 stack, or to do MLAG, or tapping, or anything like that. It creates an ecosystem similar to what we had with servers 20 years ago, where I've got an operating system that basically keeps the box running. Then I've got applications which are really the magic on top of it. That's sort of the context. What we donated was that base OS. We've worked on something called OS 10. We have an open edition of it which you can go out to the web and download for free today and start playing around with it. It's an unmodified Linux kernel currently based on the Debian distribution which we believe will serve as a solid foundation for that evolving network and ecosystem going forward. Linux foundation agreed with that and accepted our donation of that to be the foundation of the open switch project, which was talked a little bit about at this particular summit as well. We couldn't be happier to be working with the Linux foundation on the open switch project. Look forward to getting even more of the ecosystem working on that with someone like the Linux foundation behind it to build a very, very capable stack which ultimately benefits all of our customers at the end. >> Where will we see this code go into? What types of products and what markets? Is it NFE for telecom? Is it cloud servers? Where are we going to see this stuff? >> The wonderful thing about it is the answer is all of the above. That's the flexibility of it. Think of it as this way, which is maybe you have a telecom network that's focused on something like MPLS. A company that has a lot of good IP around MPLS can then write an application that can run on that base operating system giving the customer the ability to pick that specific application without having to worry about dragging on an operating system and hardware that may not be what they want. That's the telecom use case. Maybe it's a big cloud provider that has some very specific needs around an L2, L3 stack. Maybe they even have their own IP around that that they want to build on top of this OS. We've really opened up the degrees of freedom in that space across all of those industries. I certainly think where we see the early adopters and starting from that OS 10 base solution today, will be more in the telco service provider and in the cloud space, just because of the level of scale and what it is that they can benefit out of this level of flexibility. >> Excellent. >> There had to be some detractors before you open sourced this. I'm just curious, the conversation in the room about should we or should we not open source this project and take it out to the Linux foundation? What was ultimately the decision that pushed it out the door? >> Yeah, we had been working with some other open source based projects for a couple of years already, so there was a comfort level internally. But what we saw, I think going on in the networking space, was heavily influenced by what we saw going on in the server space 20 years ago when client server hit the scene. That stack became massively disaggregated. The folks who tried to keep these things stitched together into monolithic silos ultimately weren't successful. Either had to change their strategy, or drifted off into the sunset. We certainly was influenced by that history and looking forward at what we saw happening in this space. I'd say as well looking at a lot of the innovation coming out from open source projects and start ups in this space as well, doing some new and exciting things in networking. There was a big keynote yesterday and the panel discussion where a venture capitalist starting talking about, hey networking's cool again. I couldn't agree more where we're seeing startups come in and do really interesting things really, really well. What we're trying to do is create a model where those startups don't have to develop their own operating system and develop their own hardware and then all the management tools that go on top of it. Let them focus at what they're good at, which is a certain piece of IP. Let us work through things like the open networking foundation to drive disaggregation of the stack, making them successful. >> It's an interesting way too to build your community almost indirectly if you will. It's not like you have to sign a bunch of partner agreements and you can't keep track of all these startups and all your alliance people running around. But by putting it into the open source, especially with the Linux foundation just automatically, you're exposed to all these different types of new companies and innovations and that exposure goes both ways. >> Drew: Absolutely. >> It's a really cool trend, where we're seeing these big companies donate parts of these things into this formal situation that is the Linux foundation so it has a home and a place to live and grow. >> Absolutely. >> I want to shift gears a little bit. Today's keynote is about 5G. A lot of talk about 5G, mobile world congress is all about 5G but some people saying wait, it's not here yet, it's far out. But clearly, I think the message this morning from Sandra and also on the Cube yesterday is it's coming, but you don't just turn it on one day. You got to put all the pieces in place. What's Dell EMC's perspective on 5G? Where are you guys on this journey? >> For us in terms of where we play at an infrastructure level in the data center, for us, the key step right now is to get to this model where we can decouple function from location. Which is what the telecoms and the mobile networks have been trying to do through things like NFV. What we've been trying to do is help them on that journey long before we even get to the point where 5G is knocking at the door. Working with them today to put in the right infrastructure capable of supporting highly virtualized workloads and also capable of supporting a variety of different software defined networking solutions. If you get those components right, you're setting yourself up with a really good foundation for 5G. If 5G gets here and you haven't decoupled function and location yet in terms of your infrastructure or strategy that's going to be a tough one. What we're trying to do is shepherd that along and move that as fast as we can right now. >> We got Dell EMC World coming up pretty soon right? >> That's right, I hope to see you guys there. >> Previews of this? What can we expect to see? >> It will be interesting. This is the first time that as a combined company we're doing this event in Vegas. We had a preview in October as a newly closed transaction. This will have the full force effect of the combined Dell EMC firm coming together to put on a great show. Looking forward to it. Huge venue, I know you guys play a prominent role there. I'm hoping to see you guys there as well. Yes, there will be lots of announcements. I'm not going to go get myself in trouble by saying what any of those are four weeks in advance of when that's going to happen. >> No hints or anything. >> No hints, but certainly on the networking side, you'll hear a couple of announcements from us in terms of new products that we'll be talking about and stay tuned. >> I'll ask you the softer way to get to the same answer, but I know you're not going to give me the answer, but looking forward, 2017 what are some of your priorities top of mind that you guys are working on where if we see you a year from now, you'll report back that here's what we did in 2017? >> Clearly, this OS 10 strategy that we have, building upon this base is going to be key for that. Continuing to support the donations that we've made through the Linux foundation and Open Switch. Bringing in additional partners to develop on top of that to get their IP ready to be able to take advantage of that will be a key focus for us. But as well, there's some key networking speed transitions coming up that you got to keep pace with from a road map perspective, so you'll probably hear some things about that. Then as well just thinking from a Dell EMC perspective, as we look at how our portfolio as a company is evolving, a big shift toward software defined storage models, converged infrastructure, hyperconverged infrastructure. On the networking side, we're clearly trying to do everything we can to position ourselves to be a value add in any of those solutions today. That'll be the hint of the areas you can expect to hear about in May. >> That's good, that's a good little hint. It's a month to the Dell EMC World again the first combined Dell EMC World >> Drew: In Vegas. >> Well, last year we had EMC World in Vegas and Dell EMC World, it got very confusing. Now there's just one. We're like is it the Vegas one or the Austin one? So now there's just one, it's easier to keep track. >> Drew: One forum to rule them all. >> We look forward to Michael coming on, we had him on at both as well as VM World and it's always great to get his take as well. So Drew, thanks for stopping by and we look forward to seeing you in about a month in Vegas. >> Likewise, thanks guys, great time. >> Drew Schulke, Scott Reynovitch, Jeff Rick. You're watching the Cube from Open Networking Summit 2017. Thanks for watching. We'll be back after this short break. (bright music)