(upbeat music) >> Announcer: TheCUBE presents KubeCon and CloudNativeCon Europe, 2022. Brought to you by Red Hat, the Cloud Native Computing Foundation and its ecosystem partners. >> Welcome to Valencia, Spain, a KubeCon, CloudNativeCon Europe, 2022. I'm your host, Keith Townsend alongside Paul Gillon, Senior Editor, Enterprise Architecture for SiliconANGLE. We are, I think at the half point way point this to be fair we've talked to a lot of folks in open source in general. What's the difference between open source communities and these closed source communities that we attend so so much? >> Well open source is just it's that it's open it's anybody can contribute. There are a set of rules that manage how your contributions are reflected in the code base. What has to be shared, what you can keep to yourself but the it's an entirely different vibe. You know, you go to a conventional conference where there's a lot of proprietary being sold and it's all about cash. It's all about money changing hands. It's all about doing the deal. And open source conferences I think are more, they're more transparent and yeah money changes hands, but it seems like the objective of the interaction is not to consummate a deal to the degree that it is at a more conventional computer conference. >> And I think that can create an uneven side effect. And we're going to talk about that a little bit with, honestly a friend of mine Alex Ellis, founder of OpenFaaS. Alex welcome back to the program. >> Thank you, good to see Keith. >> So how long you've been doing OpenFaaS? >> Well, I first had this idea that serverless and function should be run on your own hardware back in 2016. >> Wow and I remember seeing you at DockerCon EU, was that in 2017? >> Yeah, I think that's when we first met and Simon Foskett took us out to dinner and we got chatting. And I just remember you went back to your hotel room after the presentation. You just had your iPhone out and your headphones you were talking about how you tried to OpenWhisk and really struggled with it and OpenFaaS sort of got you where you needed to be to sort of get some value out of the solution. >> And I think that's the magic of these open source communities in open source conferences that you can try stuff, you can struggle with it, come to a conference either get some advice or go in another direction and try something like a OpenFaaS. But we're going to talk about the business perspective. >> Yeah. >> Give us some, like give us some hero numbers from the project. What types of organizations are using OpenFaaS and what are like the download and stars all those, the ways you guys measure project success. >> So there's a few ways that you hear this talked about at KubeCon specifically. And one of the metrics that you hear the most often is GitHub stars. Now a GitHub star means that somebody with their laptop like yourself has heard of a project or seen it on their phone and clicked a button that's it. There's not really an indication of adoption but of interest. And that might be fleeting and a blog post you might publish you might bump that up by 2000. And so OpenFaaS quite quickly got a lot of stars which encouraged me to go on and do more with it. And it's now just crossed 30,000 across the whole organization of about 40 different open source repositories. >> Wow that is a number. >> Now you are in ecosystem where Knative is also taken off. And can you distinguish your approach to serverless or FaaS to Knatives? >> Yes so, Knative isn't an approach to FaaS. That's simply put and if you listen to Aikas Ville from the Knative project, he was working inside Google and wished that Kubernetes would do a little bit more than what it did. And so he started an initiative with some others to start bringing more abstractions like Auto Scaling, revision management so he can have two versions of code and and shift traffic around. And that's really what they're trying to do is add onto Kubernetes and make it do some of the things that a platform might do. Now OpenFaaS started from a different angle and frankly, two years earlier. >> There was no Kubernetes when you started it. >> It kind of led in the space and and built out that ecosystem. So the idea was, I was working with Lambda and AWS Alexa skills. I wanted to run them on my own hardware and I couldn't. And so OpenFaaS from the beginning started from that developer experience of here's my code, run it for me. Knative is a set of extensions that may be a building block but you're still pretty much working with Kubernetes. We get calls come through. And actually recently I can't tell you who they are but there's a very large telecommunications provider in the US that was using OpenFaaS, like yourself heard of Knative and in the hype they switched. And then they switched back again recently to OpenFaaS and they've come to us for quite a large commercial deal. >> So did they find Knative to be more restrictive? >> No, it's the opposite. It's a lot less opinionated. It's more like building blocks and you are dealing with a lot more detail. It's a much bigger system to manage, but don't get me wrong. I mean the guys are very friendly. They have their sort of use cases that they pursue. Google's now donated the project to CNCF. And so they're running it that way. Now it doesn't mean that there aren't FaaS on top of it. Red Hat have a serverless product VMware have one. But OpenFaaS because it owns the whole stack can get you something that's always been very lean, simple to use to the point that Keith in his hotel room installed it and was product with it in an evening without having to be a Kubernetes expert. >> And that is and if you remember back that was very anti-Kubernetes. >> Yes. >> It was not a platform I thought that was. And for some of the very same reasons, I didn't think it was very user friendly. You know, I tried open with I'm thinking what enterprise is going to try this thing, especially without the handholding and the support needed to do that. And you know, something pretty interesting that happened as I shared this with you on Twitter, I was having a briefing by a big microprocessor company, one of the big two. And they were showing me some of the work they were doing in Cloud-native and the way that they stretch test the system to show me Auto Scaling. Is that they bought up a OpenFaaS what is it? The well text that just does a bunch of, >> The cows maybe. >> Yeah the cows. That does just a bunch of texts. And it just all, and I'm like one I was amazed at is super simple app. And the second one was the reason why they discovered it was because of that simplicity is just a thing that's in your store that you can just download and test. And it was open fast. And it was this big company that you had no idea that was using >> No >> OpenFaaS. >> No. >> How prevalent is that? That you're always running into like these surprises of who's using the solution. >> There are a lot of top tier companies, billion dollar companies that use software that I've worked on. And it's quite common. The main issue you have with open source is you don't have like the commercial software you talked about, the relationships. They don't tell you they're using it until it breaks. And then they may come in incognito with a personal email address asking for things. What they don't want to do often is lend their brands or support you. And so it is a big challenge. However, early on, when I met you, BT, live person the University of Washington, and a bunch of other companies had told us they were using it. We were having discussions with them took them to Kubecon and did talks with them. You can go and look at them in the video player. However, when I left my job in 2019 to work on this full time I went to them and I said, you know, use it in production it's useful for you. We've done a talk, we really understand the business value of how it saves you time. I haven't got a way to fund it and it won't exist unless you help they were like sucks to be you. >> Wow that's brutal. So, okay let me get this right. I remember the story 2019, you leave your job. You say I'm going to do OpenFaaS and support this project 100% of your time. If there's no one contributing to the project from a financial perspective how do you make money? I've always pitched open source because you're the first person that I've met that ran an open source project. And I always pitched them people like you who work on it on their side time. But they're not the Knatives of the world, the SDOs, they have full time developers. Sponsored by Google and Microsoft, etc. If you're not sponsored how do you make money off of open source? >> If this is the million dollar question, really? How do you make money from something that is completely free? Where all of the value has already been captured by a company and they have no incentive to support you build a relationship or send you money in any way. >> And no one has really figured it out. Arguably Red Hat is the only one that's pulled it off. >> Well, people do refer to Red Hat and they say the Red Hat model but I think that was a one off. And we quite, we can kind of agree about that in a business. However, I eventually accepted the fact that companies don't pay for something they can get for free. It took me a very long time to get around that because you know, with open source enthusiast built a huge community around this project, almost 400 people have contributed code to it over the years. And we have had full-time people working on it on and off. And there's some people who really support it in their working hours or at home on the weekends. But no, I had to really think, right, what am I going to offer? And to begin with it would support existing customers weren't interested. They're not really customers because they're consuming it as a project. So I needed to create a product because we understand we buy products. Initially I just couldn't find the right customers. And so many times I thought about giving up, leaving it behind, my family would've supported me with that as well. And they would've known exactly why even you would've done. And so what I started to do was offer my insights as a community leader, as a maintainer to companies like we've got here. So Casting one of my customers, CSIG one of my customers, Rancher R, DigitalOcean, a lot of the vendors you see here. And I was able to get a significant amount of money by lending my expertise and writing content that gave me enough buffer to give the doctors time to realize that maybe they do need support and go a bit further into production. And over the last 12 months, we've been signing six figure deals with existing users and new users alike in enterprise. >> For support >> For support, for licensing of new features that are close source and for consulting. >> So you have proprietary extensions. Also that are sort of enterprise class. Right and then also the consulting business, the support business which is a proven business model that has worked >> Is a proven business model. What it's not a proven business model is if you work hard enough, you deserve to be rewarded. >> Mmh. >> You have to go with the system. Winter comes after autumn. Summer comes after spring and you, it's no point saying why is it like that? That's the way it is. And if you go with it, you can benefit from it. And that's what the realization I had as much as I didn't want to do it. >> So you know this community, well you know there's other project founders out here thinking about making the leap. If you're giving advice to a project founder and they're thinking about making this leap, you know quitting their job and becoming the next Alex. And I think this is the perception that the misperception out there. >> Yes. >> You're, you're well known. There's a difference between being well known and well compensated. >> Yeah. >> What advice would you give those founders >> To be. >> Before they make the leap to say you know what I'm going to do my project full time. I'm going to lean on the generosity of the community. So there are some generous people in the community. You've done some really interesting things for individual like contributions etc but that's not enough. >> So look, I mean really you have to go back to the MBA mindset. What problem are you trying to solve? Who is your target customer? What do they care about? What do they eat and drink? When do they go to sleep? You really need to know who this is for. And then customize a journey for them so that they can come to you. And you need some way initially of funneling those people in qualifying them because not everybody that comes to a student or somebody doing a PhD is not your customer. >> Right, right. >> You need to understand sales. You need to understand a lot about business but you can work it out on your way. You know, I'm testament to that. And once you have people you then need something to sell them that might meet their needs and be prepared to tell them that what you've got isn't right for them. 'cause sometimes that's the one thing that will build integrity. >> That's very hard for community leaders. It's very hard for community leaders to say, no >> Absolutely so how do you help them over that hump? I think of what you've done. >> So you have to set some boundaries because as an open source developer and maintainer you want to help everybody that's there regardless. And I think for me it was taking some of the open source features that companies used not releasing them anymore in the open source edition, putting them into the paid developing new features based on what feedback we'd had, offering support as well but also understanding what is support. What do you need to offer? You may think you need a one hour SLA for a fix probably turns out that you could sell a three day response time or one day response time. And some people would want that and see value in it. But you're not going to know until you talk to your customers. >> I want to ask you, because this has been a particular interest of mine. It seems like managed services have been kind of the lifeline for pure open source companies. Enabling these companies to maintain their open source roots, but still have a revenue stream of delivering as a service. Is that a business model option you've looked at? >> There's three business models perhaps that are prevalent. One is OpenCore, which is roughly what I'm following. >> Right. >> Then there is SaaS, which is what you understand and then there's support on pure open source. So that's more like what Rancher does. Now if you think of a company like Buoyant that produces Linkerd they do a bit of both. So they don't have any close source pieces yet but they can host it for you or you can host it and they'll support you. And so I think if there's a way that you can put your product into a SaaS that makes it easier for them to run then you know go for it. However, we've OpenFaaS, remember what is the core problem we are solving, portability So why lock into my cloud? >> Take that option off the table, go ahead. >> It's been a long journey and I've been a fan since your start. I've seen the bumps and bruises and the scars get made. If you're open source leader and you're thinking about becoming as famous as Alex, hey you can do that, you can put in all the work become famous but if you want to make a living, solve a problem, understand what people are willing to pay for that problem and go out and sell it. Valuable lessons here on theCUBE. From Valencia, Spain I'm Keith Townsend along with Paul Gillon and you're watching theCUBE the leader in high-tech coverage. (Upbeat music)

>> Announcer: Live from Las Vegas it's theCUBE covering Dell Technologies World 2018, brought to you by Dell EMC and its ecosystem partners. (bright music) >> Welcome back to Las Vegas. I'm Lisa Martin with Keith Townsend. We are with Dell Technologies World and about 14,000 other people here. You're watching theCUBE. We are excited to welcome back to theCUBE Scott Delandy, the Technical Director of Dell EMC. Hey, Scott! >> Hey guys, how are you? >> And you have a featured guest, Mike Bollman, the Director of Server and Storage Architecture from Enterprise Products Company, welcome! >> Thanks for having me. >> So you guys are a leader in oil and gas. I hear some great things. Talk to us about what it is that you're doing and how you're working with Dell EMC to be innovative in the oil and gas industry. >> So we're actually a Dell EMC storage customer for about the last two years now, and working with them on how we can bring in a lot of the data that we have from the field. The buzzword today is Internet of Things, or IoT. We've been doing it for many, many years, though, so we pull that data in and we look and analyze it and figure out how we can glean more information out of it. How can we tune our systems? As an example, one of the things that we do is we model a product as it flows through a pipeline because we're looking for bubbles. And bubbles mean friction and friction means less flow and we're all about flow. The more product we can flow the more money we can make. So it's one of the interesting things that we do with the data that we have. >> And Scott, talk to us about specifically oil and gas in terms of an industry that is helping Dell EMC really define this next generation of technology to modernize data centers and enable companies to kind of follow along the back of one and start doing IoT as well. >> Yeah, so the things that Mike has been able to accomplish within Enterprise Products is amazing because they truly are an innovator in terms of how they leverage technology to not just kind of maintain sort of the core applications that they need to support just to keep the business up and running but how they're investing in new applications, in new concepts to help further drive the business and be able to add value back into the organization. So we love working with Enterprise and users like Mike just because they really push technology, they're, again, very innovative in terms of the things that they're trying to do, and they provide us incredible feedback in terms of the things that we're doing, the things that we're looking to build and helping us understand what are the challenges that users like Mike are facing and how do we take our technology and adapt it to make sure that we're meeting his requirements. >> So unlike any other energy, oil and gas, you guys break scale. I mean you guys define scale when it comes to the amount of data and the need to analyze that data. How has this partnership allowed you to, what specifically have you guys leveraged from Dell EMC to move faster? >> So we've done a number of things. Early on when we first met with Scott and team at Dell EMC we said we're not looking to establish a traditional sales-customer relationship. We want a two-way business partnership. We want to be able to take your product, leverage it in our data centers, learn from it, provide feedback, and ask for enhancements, things that we think would make it better not only for us but for other customers. So one of the examples if I can talk to it. >> Scott: Please. >> One of the examples was early on when PowerMax was kind of going through its development cycle, there was talk about introducing data deduplication. And one of the things that we knew from experiences is that there are some workloads that may not do well with data dedup, and so we wanted some control over that versus some of the competitor arrays that just say everything's data dedup, good, bad, or indifferent, right? And we have some of that anecdotal knowledge. So that was a feature that the team listened to and introduced into the product. >> Yeah, yeah, I mean it was great because we were able to take the feedback and because we worked so closely with the engineering teams and because we really value the things that Mike brings to the table in terms of how he wants to adopt the technology and the things that he wants to support from a functionality perspective, we were able to basically build that into the product. So the technology that we literally announced earlier this morning, there are pieces of code that were specifically written into that system based on some of the comments that Mike had provided a year plus ago when we were in the initial phases of development. >> So being an early adopter and knowing that you were going to have this opportunity to collaborate and really establish this symbiotic relationship that allows you to test things, allows Dell EMC to get that information to make the product better, what is it that your company saw in Dell EMC to go, "Yeah, we're not afraid to send them back," or, "Let's try this together and be that leading edge"? >> I think honestly it came down to the very first meeting that we had. We had a relationship with some of the executives inside of EMC from other business relationships years ago, and we reached out and said, "Look, we want to have a conversation," and we literally put together a kind of a bullet-pointed list of here's how we want to conduct business and here's what we want to talk about. And they brought down some of their best and brightest within the engineering organization to have a open discussion with us. And really we're very open and honest with what we were trying to accomplish and how they could fit in, and then, again, we had that two-way dialogue back of, "Okay, well what about this," or, "What about that?" And so from day one it has been truly a two-way partnership. >> So Lisa's all about relationships and governance. I'm all about speeds and feeds. (Mike laughing) I'm a geek, and I want to hear some numbers, man. (Mike laughing) So you guys got the PowerMax. We had Caitlin Gordon on earlier. She's Product Marketing for the PowerMax, very, very proud of the product, but you're a customer that had it in your data center. Tell us the truth. (Mike laughing) How is, is it... Is it what you need to move forward? >> It is unbelievably fast in all honesty. So early on we brought it into our lab environment and we got it online and we stood it up, and so we were basically generating simulated workloads, right? And so you've got all of these basically host machines that are just clobbering it as fast as you can. We ran into a point where we just didn't have any more hardware to throw at it. The box just kept going, and it's like okay, well we're measuring 700,000 IOPS, it's not breaking a sweat. It's submillisecond (laughs) leads. It's like well, what else do we have? (laughs) And so it just became one of those things. Well, all right, let's start throwing snapshots at it and let's do this and let's do that. It truly is a remarkable box. And keep in mind we had the smallest configurable system you could get. We had the what is now, I guess, the PowerMax 2000, >> The 2000, yeah. yeah, in a very, very small baseline configuration. And it was just phenomenal in what it could do. >> So I would love to hear a little bit more about that. When we look at things such as the VMAX, incredible platform which had been positioned as a data center consolidator, but a lot of customers I saw using that as purpose-built for a mission critical set of applications, subset of applications in the data center. Sounds like the PowerMax, an example of the beta relationship you guys have, is a true platform that you can run an entire data center on and realistically get mission critical support out of a single platform. >> Absolutely, yeah, so even today in our production data center we have VMAX 450, VMAX 950s in today running. And we have everything from Oracle databases, SQL databases, Exchange, various workloads, a tremendous number of virtual iServers running on there, I mean hundreds and hundreds or actually probably several thousand. And it doesn't matter how we mix and match those. I have Exchange running on one array along with an Oracle database and several dozen SQL databases and hundreds of VMs all on one array and it's no problems whatsoever. There's no competition for I/O or any latency issues that are happening. It just works really well. >> And I think one of the other powerful use cases, if I could just talk to this, in your environment specifically there's some of the things you're doing around replication where you're doing multi-site replication, and on a regular basis you're doing failover, recovery, failback as part of the testing process. >> Mike: Absolutely. >> So it's not just running the I/O and getting the performance of the system, it's making sure that from a service-level perspective from the way the data's being protected being able to have the right recovery time objectives, recovery point objectives for all of the applications that you're running in your environment, to be able to have the infrastructure in place that could support that. >> Lisa: So I want to, oh. >> Go head. >> Sorry, thanks Keith. So I want to, I'm going to go ahead and go back up a little bit. >> Mike: Sure. >> One of the announcements that came out today from Dell Technologies was about modernizing the data center. You've just given us a great overview of what you're doing at the technical level. Where are you in developing a modern data center? Are you where you want to be? What's next steps for that? >> So I don't think we're ever where we want to be. There's always something else so we're always chasing things. But where we are today is that there's a lot of talk for the last several years around cloud, cloud this, cloud that. Everybody has a hardware, software, or service offering that's cloud-something. We look at cloud more as an operational model. And so we're looking at how can we streamline our internal business taking advantages of, say, RESTful APIs that are in PowerMax and basically automating end to end from a provisioning or a request perspective all the way through the provisioning all the way to final deployment and basically pulling the people out of that, the touchpoints, trying to streamline our operations, make them more efficient. It's been long said that we can't get more people in IT. It's just do more for less and that's not stopping. >> And if I could just make another plug for Mike, so I visited Mike in his data center it was about a year ago or something like that. And I've been in a lot of data centers and I've seen all kinds of organizations of all different size and scale and still today I talk about the lab tour that we went on because just the efficiency in how everything was racked, how everything was labeled, there was no empty boxes scattered around. Just the operational efficiency that you've built into the organization is, and it's part of the culture there. That's what gives Mike the ability to do the types of things that he's able to do with what's really a pretty limited staff of resources that support all of those different applications. So it's incredibly impressive not just in terms of what Mike has been able to do in terms of the technology piece but just kind of the people and the operational side of things. It's really, really impressive. I would call it a gold standard (Mike laughing) from an IT organization. >> And you're not biased about it. (Lisa and Mike laughing) >> Mental note, complete opposite of any data center I've ever met. (Lisa, Mike, and Scott laughing) Okay, so Mike, talk to us about this automation piece. We hear a lot about the first step to modernization is automation, but when I look at the traditional data center and I look at all the things that could be automated how do you guys prioritize where to go first? >> So we look at it from where are we spending our time, so it's really kind of simple of looking at what are your trouble tickets and what are your change control processes or trouble control tickets that are coming in and where are you spending the bulk of your time. And it's all about bang for the buck. So you want to do the things that you're going to get the biggest payback on first and then the low-hanging fruit, and then you go back and you tweak further and further from there. So from our perspective we did an analysis internally and we found that we spent a lot of time doing basic provisioning. We get a tremendous number of requests from our end users, from our app devs and from our DBAs. They're saying, "Hey, I need 10 new servers by Monday," and it's Friday afternoon, that sort of request. And so we spend the time jumping through hoops. It was like, well, why? We can do better than that. We should do better than that. >> So PowerMax built in modern times for the modern data center. Have you guys seen advantages for this modern platform for automation? Have you looked at it and been like, "Oh, you know what? "We love that Dell EMC took this angle "towards building this product "because they had the modern data center in mind"? >> So again I think it goes back to largely around REST APIs. So with PowerMax OS 5978 there's been further enhancements there. So pretty much anything that you could do before with SIM CLI or through the GUI has now been exposed to the REST API and everybody in the industry's kind of moving that way whether you're talking about a storage platform or a server platform, even some of the networking vendors. I had a meeting earlier today and they're moving that way as well. It's like whoa, have you seen what we're doing with REST? So from an infrastructure standpoint, from a plumbing perspective, that's really what we're looking at in tracking-- >> And if I can add to that I think one of the other sort of core enablers for that is just simply to move to an all flash-based system because in the world of spinning drives, mechanical systems, hybrid systems, an awful lot of administrative time is spent in kind of performance tuning. How do I shave off milliseconds of response time? How do I minimize those response time peaks during different parts of the day? And when you move to the all flash there's obviously a boost in terms of performance. But it's not just the performance, it's the predictability of that performance and not having to go in and figure out okay, what happened Tuesday night between four and six that caused this application to go from here to here? What do we have to do to go and run the analysis to figure all of that out? You don't see that type of behavior anymore. >> Yeah, it's that indirect operational savings. So before when flash drives kind of first got introduced to the market we had these great things like FaaS where you could go in and you could tune stuff and these algorithms that would watch those workloads and make their best guesses at what data to move when and where. Without flash, that's out the window. There's no more coming in on Monday and all of a sudden then something got tuned over the weekend down to a lower tier storage and it's too slow for the performance requirements Monday morning. That problem's gone. >> And when you look under the covers of the PowerMax we talked a lot today about some of the machine learning and the predictive analytics that are built into that system that help people like Mike to be able to consolidate hundreds, thousands of applications onto this single system. But now to have to go in and worry about how do I tune, how do I optimize not just based on a runtime of applications but real-time changes that are happening into those workloads and the system being able to automatically adjust and to be able to do the right thing to be able to maintain the level of performance that they require from that environment. >> Last question, Scott, we just have a few seconds left. Looking at oil and gas and what Mike and team have done in early adoption context, helping Dell EMC evolve this technology, what are some of the other industries that you see that can really benefit from this early adopter in-- >> I, what I would say is there are lots of industries out there that we work with and they all have sort of unique challenges and requirements for the types of things that they're trying to do to support their businesses. What I would say, the real thing is to be able to build the relationships and to have the trust so that when they're asking for something on our side we're understanding what that requirement and if there are things that we can do to help that we can have that conversation. But if there are things that we can't control or if there are things that are very, very specific to a small set of customers but require huge investments in terms of R&D and resources to do the development, we can have that honest conversation and say, "Hey Mike, it's a really good idea "and we understand how it helps you here, "but we're still a business. "We still have to make money." So we can do some things but we have to be realistic in terms of being able to balance helping Mike but still being able to run a business. >> Sure, and I wish we had more time to keep going, but thanks, guys, for stopping by, talking about how Dell EMC and Enterprise Products Company are collaborating and all of the anticipated benefits that will no doubt proliferate among industries. We want to thank you for watching theCUBE. I'm Lisa Martin with Keith Townsend. We're live, day two of Dell Technologies World in Vegas. Stick around, we'll be right back after a short break. (bright music)

(busy music) >> Announcer: Live from the Computer History Museum in Mountain View, California, it's the Cube, covering DevNet Create 2018. Brought to you by Cisco. (busy music) >> Welcome back, everyone. Live here in Mountain View, California, at the Computer History Museum, the heart of Silicon Valley as the Cube covering Cisco DevNet Create. I'm here with Lauren Cooney. Our next guest is Jim Bugwadia, who's the founder of Nirmata, Cube alum, here last year at our inaugural coverage of DevNet Create. Multiyear, back to back, welcome back. >> That's right, thank you, John. >> Good to see you. >> Hi, Lauren. >> All right, so last year, we kind of talked about multicloud, I think I just did a word search of the transcript last year, 13 times we mentioned multicloud. They had, Mark Zuckerberg yesterday mentioned AI 20 times in his Senate hearing. Pretty popular. What's the update this year? What's changed in year from your standpoint? >> Certainly, the trend, as we saw, starting with containers and now with Kubernetes as an operations management platform, the movement for enterprises has start adopting multicloud, whether it's hybrid or even multiple public clouds, this continues to grow. >> I mean, we're seeing a lot more Kubernetes growth this year, a lot's changed in one year. Well, as a matter fact, that's been a full year since last year, like 10 months. But still, CNCF, KubeCon is coming up in Copenhagen, we'll be there. Just in general, the open source is kind of, essentially, ratified, defacto, Kubernetes. >> Yes, and that's been great for the community as well as for enterprises, right. Because when we had multiple orchestration platforms, several contending platforms and solutions, enterprises were staying on the sidelines trying to still decide which way to go. >> What's new with your business? Give us an update on your company. >> Yeah, absolutely. 2017 is also the year when we announced our Kubernetes focused solution, so we're completely focused on operations management off Kubernetes workloads as well as clusters. We, of course, operate multicloud, so hybrid public and private. >> What's the big trend in Kubernetes? Obviously, STO's got a lot of buzz, sidecar containers, interesting. What's your analytics and your operations management software tell you about some of the trends in Kubernetes? What's hot, what's real, what's in production? Who's playing with what features? >> Yeah, so there's runtime trends, but there's also organizational trends and patterns which are very interesting. One major shift we're seeing is, whereas in the past, it was lines of businesses, or product teams directly starting to consume cloud services or Kubernetes. Now it's IT operations, right. It's also very interesting that with Kubernetes with containers, private cloud continues to may even grow in a strong fashion. We're seeing as many private cloud deployments as public cloud deployments with Kubernetes becoming that central management play. >> I think there's a research firm, Wikibon, that predicted private cloud growth. I think we nailed that one. Everyone didn't see that coming, they thought public cloud and hybrid cloud. But I mean, private cloud is happening. That's where the action is to prepare for hybrid. >> Jim: Right. >> You see that same thing? >> Absolutely. We see two factors of them. One is, of course, in highly regulated, like we are working with the energy company, we're in their environment, they will have to obviously deploy in their private network itself. However, even enterprises where they have several workloads today, they are not moving away from private cloud, but they're using public cloud for new workloads for new applications. >> They shift some of their, I won't say baggage, but their less core workloads to the cloud, analytics, other things. Let's, so on multicloud. Obviously, we debated it last year. I've been debating it all year since. I'm not, I mean, I'm bullish on multicloud in the sense it's choice. But the notion of multicloud, to me, doesn't yet exist. I mean, having apps on Azure that run on Azure and having different apps that run on Amazon, that's multiple clouds, that's not multicloud. >> Sure. >> We're starting to see some movement when people starting to think about a day layer, control planing stuff. Where are we with multicloud apps that are moving workloads? What's your analysis? >> Before we get to the application layer for multicloud, there's also the software infrastructure services that need to become multicloud. That's a lot of what we're doing at Nirmata, right. >> John: Like what, like what? >> Like, so Nirmata itself has a common management plane, a common control plane across several Kubernetes clusters, whether those are running public cloud or private cloud, and creating a common set of policies. That, in our opinion-- >> John: For infrastructure or for the apps? >> For both the clusters as well as workloads which go into the clusters. Because certainly, even if you take a containerized Kubernetes app, how you run it in production may be very different than how you run it in dev test, right. Something has to govern those policies and make sure that each cluster is set up in the right manner, so those infrastructure services first need to exist. I think the application side of things will come. We're seeing still great, a lot of innovation in the storage space. That is still a problem that needs to be solved. >> Foundationally, you agree that what you guys are working on is, foundationally, get the clusters, handle the infrastructure, get that right. >> Exactly. >> That's what's going to be dynamic. >> Absolutely. >> Don't worry about the apps yet, that kind of thing. >> Yeah, and having the ability, like one of the demos we show is the ability to take an application and to be able to create a like application in a different cloud. Now, it may not be, we might not migrate the storage, because, in production, that's not something, that's not a realistic use case. But you still want to be having the choice of being able to choose where that application gets deployed, that's a huge benefit. >> Let's go put our IT ops hat on for a second, just throw something at you. I'm an IT ops guy, I'm like, hm, we've got some on prem, we've got some Azure, I got some Amazon. I got EC2, S3, and a bunch of other stuff on Amazon, I got Azure, I forget what they use for storage, not S3, that's Amazon. Then in house, I'm running all my own provision stuff. What the heck, do I have to hire three guys? What's the, where's that come together? People get stuck there. >> Sure. Yeah, so obviously, if you're using multiple cloud providers and multiple systems, you will need some skill set, some expertise there. But more and more, the abstractions that are, again, created by Kubernetes and then with software like Nirmata is decoupling applications from that, right. It's that clean decoupling, something which we've always wanted in this space of infrastructure from applications, that's finally happening and that's really exciting. >> It'll be great when we see Office 365 running on Amazon. We made it, multicloud. >> Or at least on Linux, right, so that will happen. >> Yeah, cool. >> Great, well, when you look at these applications that you're decoupling, and I fully believe in a loosely coupled environment as well, what about the data that they can actually pull from the network? What is valuable that you kind of want to build into that application? >> Yeah, so certainly the types of data that you would, and there's systems of record and sort of systems of interaction, right. The type of data that you would probably want to keep towards your private cloud is still those systems of records, because of regulation, because of other types of requirements. But so, the engagement data, that that can be shared, distributed using some of the more innovative sort of storage concepts, distributed storage, across these clouds. >> Great, so when you're working, you've got a set of customers that, you're doing pretty well. Are you finding that, you're working with these customers that are still kind of in the old IT age, and you're kind of bringing them up to speed? Or do these guys, do they get it and they're looking for your help to really get there further and faster? >> Yeah, so when we started and Nirmata was founded in late 2013, so a lot of the conversations we had back then were why containers, why microservices. Those terms didn't even exist at that time, right. Well, containers did, but in a different form. But now it's more, enterprises know they want to go towards containers, they want to use Kubernetes, and they're looking for help and guidance in how to get there. The conversations are very different than the other. Another major trend we're seeing, like I mentioned, is the centralization of that function. Because larger enterprises are realizing that doing this in a distributed fashion, having each team build their own expertise with every cloud provider is just not scalable or cost effective, yeah. >> What's your definition of serverless? I mean, this is like a hot trend. Lambda's got functions to service. Really interesting. >> People are driving to it. >> Jim: Yeah, absolutely. >> What's your, how would you define serverless for the folks that you talk to, that say, "What is serverless?" >> Yeah, so there, one definition, obviously, the popular definition is where developers don't have to worry about the servers or any of the infrastructure, right. They're providing a function and then somehow, magically, the rest just happens, right. But I'm a software developer, I come from a development background. In any programming language, we have an object oriented, we have had lambda functions in Java as a programming language. But Java is also object oriented. My belief, and what we feel is going to happen is, applications are going to be a mix of things like serverless or lambda style functions, and stateful functions and stateless services. You need all of these in an enterprise application. It's not one or the other. >> What's the glue layer in this? >> Yeah, so that's where we feel, again, Kubernetes is the right choice. You see things like open FaaS being built on Kubernetes. We completely believe in the mission of the CNCF and how things are rolling out there, and the fact that even technologies like serverless have to start becoming decoupled from a particular provider or vendor and more of an industry standard. >> Here's a question for you. If someone asks you, hey, how should I look at the big cloud providers? Got Amazon and Google, Microsoft, you got Oracle, IBM, Alibaba, certainly for China, you go to Alibaba, they're going to cut you a deal. But I have to make some decisions about what's going to happen in the next five years as setting the foundation for an architecture. As a software engineer, what's your advice on that just getting the playbook ready, thinking about the first few steps to take, to start thinking about, OK, I'm going to be dealing with multicloud, assuming some things happen that we see connecting the dots. What's your advice? >> Yeah, so the first question is, for your business, and all of this has to be driven by business needs requirements, right, is, is it OK to be locked in into a single stack, a single vendor or provider? In a lot of cases, if you're a startup, if you're five people starting out, that may be a very good choice, and that is the most optimal path, right. Maybe you do that for a first couple of years. But as you grow, if you're an enterprise with several different teams, several applications, and if your business requires you to run on different platforms, you're going to make some different choices. That's when you would want your applications to be portable, so definitely, it makes sense to leverage cloud providers for infrastructure services, but locking in your applications to a single provider has to be carefully thought about and driven from a business perspective. >> John: You got to have choice. >> Yeah. >> All right, what's up with Nirmata? What's your action this year? What are you guys looking at doing? What's the next step? I see CNCF is doing great, good bet you guys are making. What's the product roadmap look like? What's some of the value propositions? How is this evolving, how is it evolving? >> Yeah, certainly what we see in the space, and what we're excited about, we're growing and in turn with our customers, of course. At this point, we're well funded. We're looking at doubling our head count and also-- >> John: How much did you guys raise? >> We haven't publicly announced that, but we're-- >> John: OK, but you have venture capital? >> Yeah, for the next 18 months, we're kind of funded. >> OK, a good runway. >> Yeah. Certainly, we have access to more-- >> Do you have customers? >> Yes. >> How many customers do you have? >> There's approaching the dozens now. But what we're doing-- >> Good sized customers? >> Yeah. We are focused on mid to large enterprises, right? What we do is, our appeal is to IT operations teams who are looking at deploying Kubernetes as a service for their business. As they, and IT is now sort of, in some ways, taking on this function of being able to leverage multiple cloud providers, choose where workloads go, and manage the efficiencies, manage these in their deployments. >> I mean, ops is not going away. Everything's ops now. >> Yeah, right. >> Well, thanks for coming on, appreciate it. You can see next update. >> Absolutely. >> Let me see you around KubeCon or some of the CNCF events. Great to see you. The Cube coverage here in Mountain View California for the cloud native DevOps community, part of Cisco's new foray into DevOps, DevNet Create is a part of DevNet, but this is an extension to the DevNet core Cisco Developer Conference. I'm John Furrier, Lauren Cooney, be back with more after this short break. (busy music)

(upbeat music) >> Narrator: Live from Los Angeles, it's theCUBE, covering Open Source Summit North America 2017, brought to you by The Linux Foundation and Red Hat. >> Hello there and welcome to our special exclusive SiliconANGLE Media CUBE coverage here in Los Angeles, California, for the Open Source Summit North America. I'm John Furrier, my co-host, Stu Miniman, for the two days of wall-to-wall coverage, this is day two of our coverage of what's going on in the Open Source world as the Linux Foundation consolidates their shows into a big tent event. This is the inaugural event of now consolidated, a bunch of little shows come together. This is the big show where the Linux Foundation brings their entire communities together to talk and cross-pollinate with Jim Zemlin, the Executive Director, as they outline that. Stu, we're in our kickoff of day two, we're going to do some analysis and commentary, but before we start I want to get your thoughts on just day one. Yesterday we had a lot of guests, a lot of activities going on at night, we kind of divided and conquered. What did you find out? >> Yeah so John you'd done some coverage of LinuxCon a couple of years ago, it's my first time coming to this show. We do a lot of Open Source coverage with theCUBE's over the years, so coming in it was like, okay, what are we going to be talking about, what's the vibe? And being a big tent event, you know, I was a little surprised to see, I mean, the conversation's the same that we've been hearing the last year. Kubernetes, kind of the big wave that's coming in, not just in Open Source, but really the conversation in cloud, and really was kind of the top issue that kind of containerization, the new way of architecting things, you know, Linux absolutely is down there underneath, and majorly important but, you know, it seems to be that rallying around everything Kubernetes. MesosCon's right next door, and we said two years ago you never would have thought that, Kubernetes, that Mesos would be saying, you know, the best place to run Kubernetes is on DCOS. You know, it was the container wars, the orchestration wars, all those things. Kubernetes really leading the charge there, and it really fed into a lot of the conversations we had here. And in our conversations, like with Christine Corbett, and in some of the keynotes this morning, really talking about the power of collaboration, community, you know, stuff like that, we were passionate about John. >> Yeah, I mean, Stu, here's my take on the big story coming out of L.A. for this event. And I think the top line story is this. The Open Source community has had so much success going in the early days and depending which generation you want to call it, you know, we're a little bit older, old school, maybe fourth generation, you can argue the point but here's the bottom line. The big story is that the Linux Foundation, Linux apps, are everywhere, it's a global standard, it is happening. And the scale of which the growth that's going to be coming is unprecedented, and I think for the first time in the history of the computer industry, you're seeing a pause. You're seeing a moment of excitement from the executive director, the Linux Foundation, the board members, and the participants in the community who are realizing, holy shit, this is going to grow very huge. And Open Source is going to go to a whole 'nother growth level, it's going to be exponential in scale, and you're going to see some blitzscaling going on, as Jerry Chen at Greylock and Reid Hoffman talk about. And that's going to change the nature of the participation. You're going to start to see new accelerated things, certification, the role of the foundation certainly has always been to serve the sustainable communities of Open Source. Their role will change as stewards of Open Source, the responsibility and the reliance on the Open Source software will continue to grow, and I think that scale phenomenon of Open Source is, potentially, might be the biggest wave of all, Stu, and I think some people are going to be washed like driftwood and some people are going to thrive and survive. >> You know, it's interesting, we look back at Linux, and Linux took a long time, you know, more than a decade, to really kind of gain mainstream adoption. You know, Red Hat, of course came out of with kind of the leadership and the dollars, but Linux was the foundation for everything being built today. There would be no Google without Linux. There would be no Amazon as we know it today without Linux. And I really liked, I think it was strong resonance, everybody's a little surprised, Joseph Gordon-Levitt in the keynote this morning, someone that we know, you know, from the movies, and we're here in L.A., they're like, oh great, they brought an actor. Well, he's actually pretty passionate. He has this website hitrecord.org, where they do, you know, collaboration, and it's people that are drawing and creating music and creating little clips and everything and they said how a community can help build on what they're doing. He said it's about community, fair compensation, and collaborating, rather than just socializing and sharing or any of those things. And something we've talked a lot is, what is the translation of participating in the community translate into dollars, translate into value. I know it's something you're really passionate about. >> Yeah, Stu, this is again, the big story is the growth. But let's unpack that a little bit. Open Source has always been about sharing, it's always been about community, it's been about innovation, freedom, they called them radicals in the early days but now they got to grow, flexibility, and execution. Here's the bottom line. The leadership of the Open Source is going to morph radically. Look at the program here. You got inclusion, you got a little politics, not like politics of open source, politics of cultural shaping with Christine Corbett Morgan, so she's talking about that, it's very relevant. You have Dan Lyons coming in, talking about the programmer culture, you have the actor coming in talking about collective intelligence. I believe that there's going to be a new way of how people are going to be compensated, how participation's going to scale and this comes down to some key tell tale signs. One, a new generation's coming into the Open Source world, this younger generation. They love Serverless, the love DevOps, because they don't want to deal with the infrastructure. So all the old folks, guys like our age, and gals, they have to provide leadership. I talked with Sam Ramji about this in detail, about how some of these stewards in the community have to step up and be leaders in a new way of governing because as the onboarding of more source code, more projects with IOT, with cloud, you're going to see a new generation of young developers that quite frankly are going to want to run fast, run faster, and they don't want to deal with networking, they don't want to, they want serverless, they want true programmable infrastructure, and that's going to potentially cause some changes, maybe at the leadership level but also how they run things. So, I think, Stu, this is something that we're watching as a big wave. >> Yeah, and it's funny because, we always talk, I'd love to be able to extract a way, even virtualization, oh, we're going to make it real simple, you don't have to worry it anymore, well, you know, John, we got some more interviews today, you know. Networking, storage, these things just don't magically, fairy dust, everything works really well, you know. Data has gravity, networking has lots of challenges we have to worry about. Open Source is now infused into all of these environments. Really helping to build those distributed architectures. We had a number of interviews yesterday talking about, these things are not easy, these are tough challenges. You know, even you talk to people and say, "Kubernetes is awesome," sure is not simple, it is not easy to crawl out. >> They've not graduated any projects out of the CNCF yet, talking to Chris yesterday, the COO, he said, "look, we haven't even graduated anything out of," but this is the point, Stu. Kubernetes is a tell sign, that's not fully-baked yet, it's an under-the-hood feature. Serverless, which I love the name and hate it at the same time because there's servers out there. The notion is that the due developers don't want a provision hardware, to them they just want a resource pool, so serverless is a good trend. The name is kind of weak in my opinion, but I kind of love it and hate it at the same time, I mean. >> John, it's just like cloud was 10 years ago. >> What do you think of Serverless, Stu? I mean bottom line is that how could you not like Serverless because as a developer you're just programming infrastructure as code. >> Right, absolutely, I want to be able to use things in a much more granular format, I want to be able to when I'm not using it not pay for it, it really fits into that environment. Something of course, with this show we're talking about is today, you say Serverless, I think AWS Lambda. The proprietary offering, how does something like Kubernetes fit into that? There's containers underneath, but there are a few different Open Source versions that functions as a service. There's Open FaaS, there's OpenWhisk, there's a couple of others, so how will I be able to take what we were liking about containers in general and Kubernetes specifically, that I can work across a number of environments to make sure that I'm not, John, I'm going to say the word, locked in, to a certain provider or a certain piece of the ecosystem. >> Well, Open Source is so robust right now. Again, 10% of the original ideas can be written in code that could be part of the 90% Open Source base code base. Jim Zemlin, the executive director called that the Code Sandwich. But the bottom line in my opinion, Stu, and you were just pointing it out is that the leadership has to scale. And I think one of the things that came up in some of my hallway conversations last night, talking to some folks who have been early on in Open Source, in the old days you had to hate someone, there was an enemy. There was Microsoft, and now they're on board. There was the big proprietary main mini-computer guys, the proprietary operating systems, they were the enemy. Who's the enemy now? The enemy is slowness, right? So, kind of the fundamental question is, Open Source doesn't have that enemy anymore, it's the standard. So the question is what is going to motivate the organizations? To me, I think it's speed. Speed is the new normal, scale is the new normal. Slowness and silos will be the enemy. >> Absolutely, John. It's something I've heard at a number of events we've been at recently, companies' number one thing is not cost, it's speed, and one of the reasons that so many companies work on, contribute to Open Source is to help them with that speed. They can't wait for the turn of the crank from the old software beast, or oh gosh, there are some chips or hardware involved in that? Open Source, I want to be able to contribute to the code, work on the code, ship it, move faster. >> And the other thing that came up yesterday, I want to get your thoughts and reaction to, is do you have a fashion model going on here? Never fight fashion, as we say, a good marketer would say. You have CNCF is very fashionable right now. But there's blogging and tackling projects that have been around for a while, like the networking piece. These are stable, great projects. They just don't have the pomp and circumstance as CNCF have. So, the balance of being trendy is an issue now for these Open Source communities. No one wants to work on a project that's boring but the relevance is important. So how do you react to that, Stu, because this is now a dynamic, it's kind of been there for a while, but now with the plethora of projects out there, are you nervous that fashion, fashionable trendy projects like CNCF, might suck all the option out of the governance? >> No, John, I mean, from a press and a marketing standpoint they get the attention, but I think that the stats really prove out, there's so many projects out there. Everybody's contributing to a lot of them, but it is something the developers should think about. We did an interview of a company, I remember years back, said, "how do you get the best people "and how do they choose what to do?" "Oh, whatever they feel is good." And I'm like, well, come on, you got to put a little bit of a business guidance on that to make sure what's going to help your business, what's going to help your career, if you're an individual contributing to this. There are plenty of options out there, both for starting new things as well as contributing to the big ones out there. And I liked what I was hearing from the Linux Foundation as to how they're going to give some governance to companies as to the health, that whole CHAOSS that they rolled out, talk about the health and the circular maintenance of things out there, but you know, so much activity. Kubernetes by no means is taking all of the attention, it just happens to be the current hotness. >> Well, there's some key under-the-hood details that are being worked on, that's the exciting part. Linux is a standard, it is powering. Most of the apps that are written are essential Linux apps if you look at the OS underneath. And again, the apps, again, the DevOps mindset is here, and now it's scaling and things like Serverless are going to be more greatness for developers, certainly as companies like Google, IBM, and others come in with real code and share and collaborate, a lot of people can participate in the greatness of Open Source, and I think that's, the future is bright for Linux and the Open Source Summit community. Stu, day two continues, live coverage here in Los Angeles. This is theCUBE, I'm John Furrier, Stu Miniman. Coverage of the Open Source Summit North America, in Los Angeles. We'll be right back with more after this short break. (upbeat music)