>> Live from Las Vegas, it's theCUBE. Covering AWS re:Invent 2018. Brought to you by Amazon Web Services, Intel, and their ecosystem partners. >> And welcome back, here on theCUBE, along with Justin Warren, I'm John Walls and now we're joined by Peter Nichol, who's the CEO of Instaclustr. Peter, good to see you this morning sir. >> Thank you very much John. Nice to meet you. >> and Todd Greene, CEO of PubNub. >> Good to see ya. >> Good morning Todd. >> Good morning. >> First off, let's just talk about what the two of you guys do or specifically what Instaclustr does and PubNub. Peter, if you would. >> Basically at a high level, what Instaclustr does is, we help customers to build applications that have to scale massively in a reliable way. Massive scale means terabytes or petabytes of data or even more. Reliability means the application has got to be up and running all of the time. The way we do that is, we focus on technologies in the data layer and we allow companies to essentially outsource the management of those technologies to us. So they can focus on building their application, which is what they do best, and we focus on taking a lot of the complexity away, which is helping to manage the technologies in the data layer. And the technologies that we focus on are basically in the area of storage, search, messaging, and analytics. Those technologies are Cassandra, for storage, Kafka, for messaging, Spark, for analytics, and Elasticsearch for search. We can manage all of those technologies, in any of the cloud providers, including AWS, and essentially this allows customers to outsource that and focus on their core business. We've got some great customers, PubNub being one of our best customers, a hot startup in Silicon Valley, and we're really proud to have them here with us today. >> So Todd, >> Thanks Peter. >> if you will, give us the PubNub story. >> PubNub is a company that provides a global network, which is infrastructure for real-time applications. What's a real-time application? When we started the company six, seven years ago, we made this realization that, the world was moving from applications that sort of requested data when they needed to, you know, you pull up social information, you wanted to see where something was, you ask a question, to ones where things were constantly moving and changing. So devices were emitting data and consuming data all the time. Uber was launching and everyone wanted to see where their taxi was now. Chat applications were getting big inside, dating apps and B to B apps, and B to C apps, and on top of that IoT was exploding and people needed a way to control devices and turn lights on and off. And all the infrastructure that existed at the time, didn't really address these real-time use cases. So these companies were building that stuff themselves. So PubNub launched this thing we call a Data Stream Network, but it effectively does three things. It allows you to connect to devices and leave an always-on connection over the internet, to deliver data bidirectionally to those devices. Real-time message signaling in under a quarter of a second, and then control that data going back and forth, so being able to provide logic. That core infrastructure, that sort of connect, deliver, control, powers everything from Peloton exercise bikes to Symphony Investor chat applications, athenahealth doctor, patient, nurse, kinds of collaboration and lots of IoT companies, from Logitech Harmony to Samsung smart refrigerators. Across the board, it turns out, our infrastructure has been the key to making these real-time experiences come alive. >> So you had this moment, and startups usually do, they have, you hope you do, they reach a tipping point, right, of success And things work great and you hit a boiling point (laughs) in a way, a few years back, to where things were working almost too well, and that's how you got in to Instaclustr. Tell us, give me that story if you would, or share that with our folks watching. >> Yeah absolutely, you know, it's funny, I was talking to someone recently at Amazon, at AWS, who said we rarely talk to a company your size that actually is doing more traffic than AWS is and we discovered we were doing more than twice as many messages, these control signals we talked about, around our network, more than twice as many as the world's global SMS traffic. We were doing close to 50 billion of these messages per day. So as you can imagine, that's not a simple infrastructure. We store that data, we process it, we route it, we do all these things and in one of our storage layers, built on Cassandra, we were really struggling with the expertise needed to scale this thing at the size that we needed to scale it. And we hit a tipping point about two years ago, when we realized we really needed help and we needed help immediately. We had a lot of outreach to a lot of companies, including the company themselves that had created Cassandra. But once we stumbled on Instaclustr, it was like, you know, the clouds parting, right? All of the sudden we had folks from Instaclustr on with us 24 by 7, helping us migrate, helping us move to a more stable and scaled infrastructure and we've had this ongoing relationship ever since. We now have them managing a lot of different uses of Cassandra within PubNub. >> Yeah, so, infrastructure is, (stammers) sorry, Instaclustr is built on all these open-source technologies you mentioned, like Cassandra and Spark and Kafka, but what made you choose those technologies? What was it that was attractive about them that said, you know what, this is what we want to base our company on? >> Customers are always basically looking for three things, and I think Todd summed it up very well in his business, it's basically all about scalability. If your business is successful, you want to be able to scale massively as you get more and more customers. The second thing is reliability, which means the applications have got to be always on, always up and running. The third thing is performance, which is all about latency and speed and feed and all that type of thing. We chose Cassandra because it is one of the most popular, highly scalable data bases. It's used by Apple and Netflix and big companies that have got millions of customers. We generally pick technologies, based on those three criteria, but we also focus on open-source only, for two reasons. Number one, open-source doesn't involve expensive license fees, so customers don't get locked in with expensive license fees and number two, open-source provides a degree of flexibility, cloud independence, so if you don't want to be locked in to a specific cloud provider, and you want to keep your options in the future, choose open-source. >> Okay, that's a pretty compelling sort of argument there and certainly I think the world has discovered that open-source is totally a thing that we should all be using. I'm old enough to remember when open-source was verboten and you shouldn't be using it and now it just seems to be everywhere. What is it about Instaclustr that makes you special though, because open-source, anyone could use it. I could go and download it >> Yep, yep. >> for free tomorrow, so maybe I could attempt to steal PubNub's customers, steal your customers away. So clearly that's not going to be possible for me to be able to do tomorrow. What is it about Instaclustr that you've invested in this company that makes you so special, that means that PubNub was able to rely on you? >> Right, so I think the main thing is, we have 100% of our focus on operations, not on developing proprietary IP, which we sell, which is the typical software model, we take the open-source software and we actually manage it for our customers. Basically what that means is, if they want to use Cassandra, they go to our website, they go to the customer portal, they choose the cloud provider they want to use, they choose the technology they want to use, what regions do they want to run in, what size is their cluster? They press a button and everything else is done behind the scenes by us. We do the provisioning, we install the software, and from that point on, we're managing it 24 by 7. So instead of, for example, PubNub having to build their own team for each one of these technologies, they can outsource it to us, we can do it much cheaper and we can get them to market much faster, if we're doing our job right. It's all about the operations. We can do it much cheaper and faster and that's our main advantage. The other advantage is we manage all of these different technologies in the data layer, which means that customers have one vendor they can go to, to manage several different technologies. It's all heavily, highly, integrated from one vendor. That's a big, rather than having five different vendors to manage five different technologies, we provide the complete platform. >> So Todd, what does this mean for you, now that you have this partner that you can rely on and that you can trust? What does that change for the business? What has that enabled you to be able to do now that you can look forward to saying, you know what, we can do this to grow our business. >> Well that's a good question. Like Instaclustr, we operate PubNub. Customers pay us, not for our technology, but for our ability to operate our technology at massive scale. And we provide five nines SLA, which is a fancy way for saying, if we have an outage for more than 26 seconds in a month, we provide credits back to our customers. That's a really hard, high bar to fill. And so philosophically, we see ourselves as an operations company ourselves, right? And so we're very careful about who we would bring in to the fold as part of operations, right? And so it has to be an organization that has the same security levels that we do, SOC 2 Type II Compliance, has the same understandings and philosophy around operating things at high availability, and can do it in a way that we feel like, you know, in many ways is a part of our team and not some vendor that we don't know how to get on the phone. Not some vendor that we don't really trust, right? It has to feel like it's part of our company. So really it's only been Instaclustr that we've been able to develop that trust around. And so it is actually in all of us to sort of focus on areas where we can do more innovation while keeping the five nines SLAs at 26 seconds minimum, you know, maximum, of issues any month, but allow us to focus a lot more on innovation and not on the things that, frankly, Instaclustr, as far as we can tell, is best in the world at, which is really operating this infrastructure, the Cassandra piece. >> And what do you want to take on then? You told about innovation. If there's an area of your business, you say alright, this is where 2019, where I want it to take us, what would that be? >> It's a great question. One of the big changes for PubNub, was that we built our initial business on the backs of other startups and it was great. We got to some level of scale by powering a lot of innovative interesting applications that were themselves trying to be the first real-time this and the first real-time that and the first real-time the other thing. And then about two years ago something happened, a year and a half ago, that need for real-time, for having things update in real time, inventory, prices, chat applications, moving things on a map, seeing where your trucks were, that went mainstream, and now even the largest app companies in the world, if they release any kind of application, whether it's a business application or a consumer app, if it doesn't have that same real-time experience like an Uber or like a Snapchat, people kind of look at it and say, well this feels like it was built 20 years ago, right? And so what's happened to our industry, has been the moving of the need for this real-time experience, into the mainstream. Now that's been great for us, but it also means as we are selling to a larger and larger group of, we call mainstream larger enterprise customers, the way we package our product, the way we make it consumable by larger companies, make it easier to deploy our product, make it easier to understand and adding features that round that out, is really the core of our focus right now. Is really being able to appeal to those larger companies. We already have the scale, in fact, we recently participated in an event which was the Guinness Book of World (stammers) Record's largest online event in history. And we powered the source in India for Cricket, we powered the largest social interaction, over 10 million people synchronously going through our network, all in one virtual environment. So we know we can scale this thing beyond any existing human need and now it's really about making sure it's accessible to the world's largest companies. >> So it was cricket in India? >> Yes, yes. >> I would've thought it was the Justin Warren fan club, but I guess not, I (stammers) second online, right? >> Yeah, probably. >> There's a lot of people in India who love cricket, and they all have mobile phones. >> Yes, well gentlemen, thanks for being with us, Peter, Todd, continued success and then thanks for being here on theCUBE. >> Okay, thank you very much. >> Thank you so much, it's been a pleasure, thank you. >> We continue live coverage here from Las Vegas. We're in the Sands. We're here all week at AWS re-Invent. (calm digital music)

>> Announcer: Live from San Francisco, it's theCUBE covering DevNet Create 2017 brought to you by Cisco. >> Hey, welcome back everyone, we are here live in San Francisco for theCUBE's exclusive two days of coverage of Cisco's new inaugural event called DevNet Create. An extension of their classical developer group DevNet, DevNet Create really going into the ethos of DevOps, the infrastructruous code targeting cloud-native and app developers, the collision between applications and infrastructure. A new direction for Cisco, this is theCUBE, I'm John Furrier with my co-host Peter Burris. Our next guest is Bhavana Srines- >> Bhavana: Srinivas. >> Srinivas, solutions architect at PubNub, which provides real-time pubs. Welcome to theCUBE, thanks for coming on. >> Thank you, it's great to be here and talk to you guys. >> So, PubNub, you couldn't get PubSub but it relates. Explain what you guys do real quick. >> Yeah, so what PubNub is is it provides real-time infrastructure as a service. So we realized that a lot of people were trying to build these real-time, always-on applications wherein when something happens in real life, you want that message or event to be translated to several of your friends or other users instantly. So, everyone is trying to build a real-time app like, say, a taxi dispatcher like Lyft or for example, a chat application where if I send a message my friends need to receive it instantly. Anyone trying to build these kind of real-time applications were building the infrastructure before they even got to the best part, which is their business logic. So, we decided that we will provide that infrastructure, we'll provide that plumbing. We'll build a global distributed network for all of these app developers to build their always-on applications. So, what we do is provide this real-time, bidirectional communication between devices in a very scalable manner and it's very, it focuses on real-time communication. >> And the key there is that most apps are mobile, you require this so you want to get them accelerated because, let's face it, most apps don't make it, right? So why build all the plumbing? >> Bhavana: Right. >> Focus on getting to figuring out the best app experience. >> Exactly, so it's for mobile, web, and even for IoT devices because everyone now wants to talk to each other. You're not going to let that gradual sit by itself, you want to connect it. So, like you said, it's meant to go to market quickly. Like you said, not every company has the resources or the time and the effort to put in to building this infrastructure, so why don't we provide this as a service? So now they're focusing on their business logic and try and make that app look pretty. >> So you're clearly in the world of cloud-native, which really is pure cloud, mostly startups. Because why have a data center? If you're a startup, I mean anyone that does a startup these days, if you have data center you're either crazy or you have so much case you just want to spend it. Why would you want to do it? You just go right to the cloud. >> Right, right, right, so we call ourselves more of a network because we're not... Think of it as a CDN but for real-time data. It's not longer static files- >> Peter: CDN is smaller messages deterministic performance. >> Exactly, exactly, you nailed it. So, what we're- >> John: You nailed it. >> Peter: I'm the man! (laughter) >> All right, so talk to me about where your use cases are. Give us some examples of customers and some specific apps that are on the network. >> For sure, so, if you take eBay for instance. They use PubNub for buyer-seller chat. So, you go on eBay, you want to talk to that buyer before you actually buy that thing. So, that chat application is powered by PubNub. Or for instance, you go to Logitech and then you want to control all these devices in your house, and PubNub is what enables that from your mobile phone to all the devices in your house. That is PubNub in play there. Or, if for instance, Lyft uses us for to see where exactly the driver is in real-time. So you're able to see every instantly. So, it's such a low-lying infrastructure that we play in almost 35 different industries. Whether it's real-time chat or taxi dispatch, multiplayer game, like Pocket Gems uses us. That's where it's real-time at it's core, right? So, you have two screens, people are playing a game. You want to see what the other person is doing, right? That's the essence of a multiplayer game. And so you can imagine how important it is to be real-time in such a use case, and that's where PubNub fits in. >> But just so we're clear, we're not talking about scada kinds of, system control kinds of things. Low-level IoT protocols, we're talking about a machine that serve human types of speeds. >> Bhavana: Exactly. >> A few hundred milliseconds, that kind of stuff. >> Exactly, exactly, so we're protocol, we call ourselves protocol agnostic. So, as long as a device can speak ECP it can understand PubNub. So, all you're dealing with is that high-level application level APIs. >> Peter: So you're still layer seven? >> Yes, very much layer seven. >> Peter: That's important. >> Yeah, but then the way we provide that layer seven API is by building out this very robust network. >> All right, so explain to us how you guys play with microservices because you're doing a topic on, Always-on Apps are Driving Microservices to the Edge. >> Yeah, yeah, yeah, yeah. So, so far you understood that PubNub's almost like a message pipeline between devices. If you have a message to throw out, PubNub will route it for you anywhere in the world. So then we decided that people are sending all these small bits of data through our network, but let's do something with that data. So, maybe if there's a chat application and people are talking to each other, maybe you want to translate it in-stream. So you put in a function there on the PubNub network that says, "Hey, if my destination is going to a "Spanish-speaking person, translate it." Or if I want to do sentiment analysis because I have a customer support kind of app, data's flowing between an agent and a customer, then let's do some sentiment analysis on it. So, what we added to this humongous network is the ability to put small pieces of logic on it. So that it acts on the data flowing though the network. And so it becomes easy to spin up these microservices through PubNub and that's what I'm going to be talking about. So it's, yeah-- >> John: So it's a brand new innovation. >> Bhavana: Sorry? >> A new innovation opportunity for you guys-- >> Bhavana: Exactly. >> To apply logic into a data stream while it's in motion. >> Of course, yes, so we recently even did have a BLOCK, we call this BLOCK, event handlers. So, we have a BLOCK with Cisco Spark. So if you wanted to do any kind of collaboration using that Cisco Spark, you can now send data through PubNub and instantly, in real-time it will sync up with Cisco Spark. >> So, Bhavana I got to get your perspective on something. We talk to a lot of enterprises and you're involved with a lot of cutting edge companies. Microservices, cloud-native, they're doing cutting edge stuff. They don't have time to be bothered with old-fashioned stuff because they have no baggage. There's no legacy, a lot of these enterprises have legacy environments, they're trying to be relevant, and they're looking to design great apps. Is there a pattern that you've seen or observation that you've noticed on these successful new, emerging companies that could be an opportunity for someone to look at and say, "Hmm, I should to more of that." What's the trend? >> (laughs) That's a loaded question. But we talk to a lot of small and medium businesses and also a lot of enterprise level companies. But then, it's just that the sales cycles are much slower. You can't go to a company and say that, "Hey, I know you're building a technical product. "Speed up your development process," right? So it's up to them to do that. So with enterprises at least they have the resource and time to do so. But, like you said, they have a lot of legacy systems. So, it's hard to tear that down and-- >> John: Build new. >> Build new stuff that you have, which might be more optimized but we try to make it work. So we're trying to now, like I said, if you're within the PubNub eco-system, you can use our BLOCKS but then everyone understands https. So we've now included a BLOCKs endpoint, where if you just dock http, you can get in to the PubNub network. So ways to use our network using their infrastructure. So we're trying to make this network accessible for anyone, irrespect of whether they use-- >> John: So integrate easily into these older legacy environments. >> Bhavana: Exactly. >> Well, but so one of the places where at least PubSub initially started was the idea that you could have a published without having to know who the clients were. >> Bhavana: Right, right, right, right. >> So you anticipate, does PubNub anticipate that you're actually going to be in a position to say, I as a real-time device, who has a real-time service, can put something into PubNub and then devices out there can subscribe to it? So a device manufacturer may sell something, it takes advantage of that centralized service, but have it operate in a deterministic, high-quality high-reliability way? Is that kind of the direction you're taking? >> Yeah, but at the end of the day, someone has to build an application. >> Peter: Sure. >> So for instance, even in Insteon, they use PubNub. They integrate PubNub within their devices and they're now selling it at Best Buys and whatnot. So it's like when I as a customer buy an Insteon product, I don't know there's PubNub in there. But then using PubNub, Insteon's now able to collect data about my usage patterns or where I can be saving energy, et cetera so the- >> But then the alternative for them is to build a full-stack system, manage it, have system integrators, have operators-- >> I mean, that was exactly the case at Insteon. They had 23 on-call support agents all day, every day, trying to do exactly what PubNub did for them without that. >> John: Yeah, they save all that cost. It's kind of like why people use Amazon. >> Right, exactly. >> (laughs) I don't need a data center, I don't need staff. All right, what did you think about this event? Obviously, Cisco has been first in a lot of markets and succeeded in networking but didn't really knock it out of the park on smart home or-- >> Peter: Linksys. >> (laughs) And so on and so forth but now, with cloud-native, we're saying is that the opportunity for them? >> Bhavana: Yeah. >> What's your take on Cisco's moving up the stack? >> I mean, I think it's great. This is one of the first conferences that DevNet is hosting for developers, right? I just got here but we've had a booth here and people are saying really great things. And there's been a steady crowd. And apparently there have been great talks. So I'm actually very excited to give my talk and then go on. >> Peter: What time is your talk, today? >> Yeah, today at 5 p.m. and then I'm here tomorrow as well. So, excited to check out the whole experience. >> Great to have you on theCube and thanks for sharing PubNub and we look forward to getting more updates from you. And congratulations on your success. >> Bhavana: Thank you. >> And your customers, thanks for sharing. >> It was great to be here, thank you so much. >> John: Thanks so much. >> So you should stop by our booth when you- >> We'll stop by and check out PubNub, the real-time pub-sub service used by all cutting edge companies in the cloud-native. This is theCUBE, Cube Cloud. Check out our content at youtube.com/siliconangle. That's our Cube Cloud, all the content there for you. I'm John Furrier with Peter Burris. Stay with us for more live, exclusive coverage from Cisco's inaugural event, DevNet Create, after this short break. (upbeat music) >> Hi, I'm April Mitchell and I'm the Senior Director of Strategy and Planning.

(energetic music) >> [Narrator] From San Francisco, it's theCUBE, covering Radio 2018. Brought to you by VMware. >> Hello everyone. Welcome to the special CUBE coverage here in San Francisco, California for VMware's Radio 2018 event. This is their R&D big event kickoff. It's like a sales kickoff for engineers, as Steve Herrod said on stage. Out next guest is Mornay Van Der Walt, VP of the Explore Group, Office of the CTO. Also, program chair of the Event Today Conference, working for the collective of people within VMware on a rigorous selection committee for a high bar here at your event. Welcome to theCUBE. Thanks for joining me. >> Thank you. >> Talk about the event, because I know a lot of work went into it. Congratulations, the talks were amazing. I see the schedule. We have Pat Gelsinger coming on later today. We just had Ray O'Farrell on. This is like the, I don't want to say, Burning Man of Vmware, but this is really a recognition, but also really important innovation. Take a minute to talk about the process that you go through to put this together. It's a fantastic event. The smartest minds, the cream rises to the top. It's hard, it's challenging, it's a team effort, but yet you gotta ride the right waves. >> Right. So, RADIO: R&D Innovation Offsite. And as you said, it is tough because we've got this huge R&D community and they've all got amazing ideas. So they get the opportunity to submit ideas. I think this this year we have over 1,700 ideas submitted, and at the end of the day we're only going to showcase 226 of those ideas across research programs, posters, breakout sessions, Just-In-Time BOFs, Birds Of a Feather. You know, so, the bar is high. we've got a finite amount of time, but what's amazing is we take these ideas, and we don't just showcase them at RADIO. We have four other programs that give us the ability to take those ideas to the next level. So when we think about the innovation programs that come out of OCTO, this is really to drive what we call "Off-Road Map Innovation." So Raghu and Rajiv, with our Product Cloud Services Division, are driving road map, zero to three years out the stuff that you can buy from sales, >> [Furrier] Customer centric? >> Customer centric, yeah. OCTO is providing an innovation program structure, these five programs: Tech Talks, Flings, Borathons, RADIO, and xLabs, and as a collective, they are focused on off-road map innovation. Maybe something that's-- >> Give me an example of what that means, Off-Road Map. >> Sure. So last year at RADIO we did a paper that was showcased on functions as a service. So you think of AWS Lambda, right. [Furrier] Yep, yep >> VM was uniquely positioned, with the substrate, to manage and orchestrate VM's containers and whynot functions. So this radio paper was submitted, I then, as the xLabs group, said we're going to fund this, but given where we are in this market, we said, "Alright, we'll fund this for 12 months." So, we're incubating functions as a service. In July/August time frame, that'll actually exit xLabs into the Cloud Native business. >> It's a real rapid innovation. >> Very rapid. >> Within a 12 month period, we're gonna get something into a BU that they can take it to market. >> Yeah, and also I would say that this also I've seen from the talks here, there's also off-road map hard problems that need to kind of get the concepts, building blocks, or architecture... >> [Van Der Walt] Correct. >> With the confluence of hitting, whatever, its IOT or whatever, blockchains, seeing things like that. >> [Van Der Walt] Yeah. Correct. >> Is that also accurate too? >> Very true. And, you know, Ray had a great slide in his keynote this morning, you know, we spoke about how we started in 2003, when he joined the company, it was all about computer virtualization. Fast-forward 15 years, and you look at our strategy today, it's any Cloud, any device, any app, right? Then, you gotta look to the future, beyond there, what we're doing today, what are the next twenty years going to look like? Obviously, there's things like, you know, blockchain, VR, edge computing, you know, AIML... >> [Furrier] Service meshes? >> Services meshes, adaptive security. And, you know, people say, "Oh, AIML, that's a hot topic right now, but if you look back at VM ware, we've been doing that since 2006. Distributed resource scheduler: a great example of something that, at the core of the product, was already using ML techniques, you know, to load-balance a data center. And now, you can load-balance across Clouds. >> It's interesting how buzzwords can become industry verticals. We saw that with Hadoop; it didn't really happen, although it became important in big data as it integrates in. I mean, I find that you guys, really from the ecosystem we look at, you guys have a really interesting challenge. You started out as "inside the box," if you will. I saw your old t-shirt there from the 14 year history you guys have been doing this event. Great collection of t-shirts behind me if you can't see it. It's really cool. But infrastructures, on premise, you buy, it's data center, growth, all that stuff happened. Cloud comes in. Big data comes in. Now you got blockchain. These are big markers now, but the intersection of all these are all kind of touching each other. >> [Van Der Walt] Correct. >> IOT...so it's really that integration. I also find that you guys do a great job of fostering innovation, and always amazed at the VM world with some great either bechmarks or labs that show the good stuff. How do you do it? Walk me through the steps because you have this Explorer program, which is working. >> [Van Der Walt] Yeah >> It's almost a ladder, or a reverse ladder. Start with tech talks, get it out to the marketplace... >> [Van Der Walt] Do a hackathon. >> Hackathon. Take us through the process. So there's four things: tech talks, borathons, which is the meaning behind the name, flings, and xLabs. >> Correct >> Take us through that progression. >> ... and RADIO, of course. >> And RADIO, of course, the big tent event. Bring it all together. >> So, I'm an engineer. I have a great idea. I wanna socialize it; I wanna get some feedback. So, at VMWare, we offer a tech talk platform. You come, you present your idea. It's live. There'll be engineers in the audience. We also record those, and then those get replayed, and engineers will say, "You know, have you thought about this?" or "Have you met up with Johnny and Mary?" They're actually working on something very similar. Why don't you go and, you know, compare ideas? I can actually make that very real. I was in India in November, and we were doing a shark tank for our xLabs incubator, and this one team presented an idea on an augmented reality desktop. We went over to another office, actually the air watch office, and we did another shark tank there. Another team pitched the exact same idea, so I looked at my host, and I said, "Do these two teams know each other?" and the guy goes, "Absolutely not," so what did we do? We made the connection point. Their ideas were virtually identical. They were 25 kilometers apart. Never met. >> [Furrier] Wow. >> You know, so when, that's one of the challenges when your company becomes so big, you've got this vast R&D organization that's truly global, in one country 25 kilometers apart, you had two teams with the same idea that had never met. So part of the challenge is also bringing these ideas together because, you know, the sum of the parts makes for a greater whole. >> And they can then collectively come together then present to RADIO one single paper or idea. >> [Van Der Walt] Absolutely, or go ahead and say, you know what, let's take this to the next step, which would be a borathon, so borathons are heckathons. >> Explain the name because borathon sounds like heckathon, so it is, but there's a meaning behind the name borathon. What is the meaning? >> Sure. So, our very first build repository was named after Bora Bora, and so we paid homage to that, and so, instead of saying a heckathon, we called it a borathon. And one of our senior engineers apparently came up with that name, and it stuck, and it's great. >> So it's got history, okay. So, borathons is like ... okay, so you do tech talks, you collaborate, you socialize the idea via verbal or presentation that gets the seeds of innovation kinda planted. Borathon is okay, lets attack it. >> Turn it into a prototype. >> Prototype. >> And it gets judged, so then you get even more feedback from your most senior engineers. In fact ... >> And there's a process for all this that you guys run? >> Yeah, so the Explorer groups run these five innovation programs. We just recently, in Palo Alto, did a theme borathon. Our fellows and PE's came together. Decided the theme should be sustainability, and we mixed it up a little bit. So, normally, at a borathon, teams come with ideas that they've already been developing. For this one, the teams had no idea what the theme was going to be, so we announced the theme. Then, they showed up on the day to learn what the five challenges were going to be, and some of those challenges, one of them was quite interesting. It was using distributed ledger to manage microgrids, and that's a ... >> A blockchain limitation >> Well, it's a project that's, you know, is near and dear to us at VMWare. We're actually going to be setting up a microgrid on campus, and if you think about microgrids, and Nicola Acutt can talk more to this, we're gonna be looking at, you know, how can we give power back to the city of Palo Alto? Well, imagine that becoming a mesh network. >> [Furrier] With token economics. >> How do you start tracking this, right? A blockchain would be a perfect way to do this, right? So, then, you take your ideas at a borathon, get them into a prototype, get some more feedback, and now you might have enough critical mass to say, "Alright, I'm going to present a RADIO paper next year." So, then, you work as a team; get that into the system. >> [Furrier] And, certainly, in India and these third-world countries now becoming large, growing middle-class, these are important technologies to build on top of, say, mobile... >> [Van Der Walt] Absolutely. >> And with solar and power coming in, it's a natural evolution, so that's good use case. Okay, so, now I do the borathon. I've got a product. Flings? >> It's a prototype, right, so now ... >> You can socialize it, you have a fling, you throw it out there, you fling it out there What happens? >> Yeah, so, I've done something at a borathon. It's like, I want to get some actual feedback from the ecosystem: our customers and partners. That example I used with vSAN. You know, vSAN launched. We wanted to get some health analytics. The release managers were doing their job. The products got a ship on the state. Senior engineers on the team got a health analytics tool out as a fling. It got incredible feedback from the community. Made it into the next release. We did the same with the HTML clients, right? And that's been in the press lately because, you know, we've got Rotoflex. Now, there's HTML, but that actually started - two teams started working on that. One team just did HTML >> a very small portion of the HTML client, presented a RADIO paper. Two years later, another team, started the work, and now we have a full-fledged HTML client that's embedded into the VIS via product. >> [Furrier] So, the fling brings in a community dynamic, it brings in new ideas, or diversity, if you will. All kinds of diverse ideas melting together. Now, xLabs, I'm assuming that's an incubator. That brings it together. What is xLabs? Is that an incubator? You fund it? What happens there? >> So with an xLabs, the real way to think about it, it's truly an incubator. I don't want to use the word "start-up" there because you've clearly got the protection of the larger VMware organization, so you're not being a scrappy start-up, but you've got a great idea, we see there's merit ... >> [Furrier] Go build a real product. >> We see it more being on the disruptive side, and so we offer two tracks in the xLabs. There's a light track, which typically runs three to six months, and you're still doing your day job. You know, so you're basically doing two jobs. You know, we fund you with a level of funding that allows you to bring on extra contracting, resources, developers, etc., and you're typically delivering one objective. The larger xLab is the full-track, so functions as a service. Full-track, we showcased it as a RADIO paper last year. We said, "Alright, we're going to fund this. We're going to give it 12 months worth of funding, and then it needs to exit into a business unit," and we got lucky with that one because we were already doing a lot of work with containers, the PKS, the pivotal. >> [Furrier] Do the people have to quit their day job, not quit their day job, but move their resource over? >> [Van Der Walt] Absolutely. >> The full-track is go for it, green light >> Yep >> Run as fast as you can, take it to this business unit. Is the business unit known as the end point in time? Is it kind of tracked there, or is it more flexible still. >> Not all the time. You know so sometimes, with functions it was easier, right? So, we know we've got pull for zone heading up Cloud native apps. The Cloud native business unit is doing all the partnerships with PKS. That one makes sense. >> [Furrier] Yeah. >> We're actually doing one right now, another xLabs full, called network slicing, and it's going to play into the Telco space. We've obviously got NFV being led by Shekar and team, but we don't know if network slicing, when it exits, and this one is probably going to have a longer time arise and probably 24-36 months. Does it go into the NFV business unit, or does it become its own business unit. >> [Furrier] That's awesome. So, you got great tracks, end to end, so you have a good process. I gotta ask you the question that's on my mind. I think everyone would look at this, and some people might look at Vmware as, and most people do, at least I do, as kind of a cutting-edge tier one company. You guys always are a great place to work. Voted as, get awards for that, but you take seriously innovation and organic growth in community and engineering. Engineering and community are two really important things. How do you bring the foster culture because engineers can be really pissed off. "Oh my god! They're idiots that make the selection!" because you don't want engineers to be pissed cuz they're proud, and they're inventing. >> Yep, yep. >> So, how to manage the team approach? What's the cultural secret in the DNA that makes this so successful over 14 years? >> So, before I answer that question, I think it's important to take a step back. So, when we think about innovation, we call this thing the Vmware "innovation engine." It's really three parts to it, right? If you think about innovation at its core: sustaining, disruptive, internal, external, And, so, we've got product Cloud Services group, Raghu and Rajiv, we've got OCTO, headed up by Ray, we've got corp dev headed up by Shekar. Think of it as it's a three-legged stool. You take one of those legs away, the stool falls over. So, it's a balancing act, right? And we need to be collaborating. >> [Furrier] And they're talking to each other all the time. >> We're talking to each other all the time, right? Build or buy? Are we gonna do something internal, or we gonna go external, right? You think something about acquisitions like Nicira, right? We didn't build that; we bought it. You think about Airwatch, right? Airwatch put us into the top right quadrant from Gartner, right? So, these are very strategic decision that get made. Petchist presented at Dell emc world, Dell Technologies world. He had a slide on there that showed, it was the Nicira acquisition, and then it sort of was this arc leading all the way up to VeloCloud, and when you saw it on one slide, it made perfect sense. As an outsider looking in, you might have thought, "Why were they doing all these things? Why was that acquisition made? But there's always a plan, and that plan involves us all talking across. >> [Furrier] Strategic plan around what to move faster on. >> Correct >> Because there's always the challenge on M&A, if they're not talking to each other, is the buy/build is, you kinda, may miss a core competency. They always ... what's the core competency of the company? And should you outsource a core competency, or should you build it internally? Sometimes, you might even accelerate that, so I think Airwatch and Nicira, I would say, was kinda on the edges of core competency, but together with the synergies ... >> [Van Der Walt] Helped us accelerate. >> And I think that's your message. >> [Van Der Walt] Yep. >> Okay, so that's the culture. How do you make, what's the secret sauce of making all this work? I mean, cuz you have to kinda create an open, collaborative, but it's competitive. >> [Van Der Walt] Absolutely. >> So how do you balance that? >> You know, so clearly, there's a ton of innovation going on within the prior Cloud services division. The stuff that's on the truck that our customers can buy today, alright? We also know we gotta look ahead, and we gotta start looking at solving problems that aren't on the truck today, alright? And, so, having these five programs and the collective is really what allows us to do that. But at the same time, we need to have open channels of communication back into corp dev as well. I can give you examples of, you know, Shekar and his team might be looking at Company X. We're doing some exploratory work, IOT, I did an ordered foray. IOT is gonna be massive; everybody knows that, but you know what's going to be even more massive is all the data at the edge, and what do you do with that data? How do you turn that data into something actionable, right? So, if you think about a jet engine on a big plane, right? When it's operating correctly, you know what all the good levels are, the metrics, the telemetry coming off it. Why do I need to collect that and throw it away? You're interested in the anomalies, right? As we start thinking about IOT, and we start thinking all this data at the edge, we're going to need a different type of analytics engine that can do real-time analytics but not looking at the norm, looking at the deviations, and report back on that, so you can take action on that, you know? So, we started identifying some companies like PubNub, Mulesoft, too, just got acquired, right? Shekar and his team were looking at the same companies, and was like, "These companies are interesting because they're starting to attack the problem in a different way. We do that at Vmware all the time. You think about Appdefense. We've taken a completely different approach to security. You know what the good state is, but if you have a deviation, attack that, you know? And then you can use things like ... >> It's re-imagining, almost flipping everything upside-down. >> Yeah, challenging the status quo. >> Yeah, great stuff, great program. I gotta ask you a final question since it's your show here. Great content program, by the way. Got the competition, got the papers, which is deep, technical coolness, but the show is great content, great event. Thanks for inviting us. What's trending? What's rising up? Have you heard or kind of point at something you see getting some buzz, that you thought might get buzz, or it didn't get buzz? What's rising of the topics of interest here? What's kind of popping out for you; what's trending if I had to a Twitter feed, not Twitter feed, but like top three trending items here. >> Well, I'll take it back to that last borathon that we did on sustainability. We set out the five challenges. The challenge that got the most attention was the blockchain microgrid. So, blockchain is definitely trending, and, you know, the challenge we have with blockchain today is it's not ready for the enterprise. So, David Tennenhouse and his research group is actually looking at how do you make blockchain enterprise ready? And that is a difficult problem to solve. So, there's a ton of interest in watching ... >> [Furrier] Well, we have an opinion. Don't use the public block chain. (both laugh) >> So, you know, that's one that's definitely trending. We have a great program called Propel, where we basically attract the brightest of the brightest, you know, new college grads coming into the company, and they actually come through OCTO first and do a sort of onboarding process. What are they interested in? They're not really interested in working for a particular BU, but, you know, when we share with them, "You're gonna have the ability to work on blockchain, AI, VR, augmented reality, distributed systems, new ways of doing analytics >> that's what attracts them. >> [Furrier] And they have the options to go test and put the toe in the water or jump in deep with xLabs. >> Absolutely >> So, I mean, this is like catnip for engineers. It draws a lot of people in. >> Absolutely, and, you know, we need to do that to be competitive in the valley. I mean, it's a very hard marketplace. >> Great place to work. >> You guys have a great engineering team. >> Congratulations for a great event, Mornay, and thanks for coming on theCUBE. We're here in San Francisco for theCUBE coverage of RADIO 2018. I'm John Furrier. Be back with more coverage after this break. Thanks for watching. (upbeat techno music)

>> Announcer: Live from San Francisco, it's theCUBE, covering DevNet Create 2017, brought to you by Cisco. >> Hello everyone, welcome to day two of theCUBE's exclusive coverage Cisco Systems' DevNet Create, their inaugural event where they're put in the foray into the developer community and the open source community, really looking at DevOps, cloud-native, with data. Great move by Cisco. We're going to analyze it again here on the opening day and review yesterday a little bit and talk about what we think is happening here and give you the take on it, our angle, and extracting the signal from the noise. I'm John Furrier with my co-host Peter Burris for day two, head of research for SiliconANGLE Media, also, general manager of Wikibon.com. Check out Wikibon.com, great research, some free, most of it. The good stuff is behind subscription firewall via client. Check it out. Your business will do great because of it. Peter, quick plug there. Get that out of the way. But, let's talk about Cisco. >> Thank you, John. >> We said yesterday and kind of played out Cisco has an opportunity to transfer and grow their core development community, DevNet, which is only three years old, is very robust, it's the heartbeat of Cisco right now, it's the core constituency for Cisco Systems. We're talking tier one elite networking guys, the plumbers, (laughs) pack it movers, whatever you want to call them, they've been designing networks from internet scale to today, everything from Voice over IP, you name it, they've been doing it and the network has been the center of the action and the data center. It's been the most critical asset for availability, operational support, and stable. But not very adaptive, not very agile (laughing) so, you know, now-- >> Almost by definition. >> And that's the purpose is the network guys drive the network and enable opportunities, but now that shift has happened. It's DevOps application developers are driving change to the network and the big conversation is what does it mean for Cisco? What does it mean for their communities? And we were saying that this is an opportunity to extend. We had the folks first from RedHat on yesterday talking about how when you have things come together, there's opportunities. There could be a collision or there could be an opportunity so, can use data science and computer science come together. That is now big data. That's changed the game in the world. Cisco now has DevOps and networks coming together. Applications and infrastructure. This is an opportunity for Cisco. Your take from yesterday, do you feel it? Do you think it's real? What's your take? >> Well, it's real. Digital transformation's happening and it's happening because people can now do things with data that they couldn't do before and they're starting to. So, that's the base of digital transformation, but the reality is, in a digital business, you're going to be by definition, almost inherently, highly networked. And your ability to move data where it needs to be when it needs to be there to whom it needs to be so that they can consume it, is emerging as an essential capability. But, you're not going to do that manually everywhere. You're not going to do it manually in the infrastructure, it's just too complex. And you're certainly not going to do it manually in the applications. And so, you're absolutely right. Being at that edge, being at that margin between how the network does things, how the network is the basis for very high-quality capability for distributing application componentry and how the data then flows over that network under control by applications is really where the next five years are going to be, a lot of the new value is going to created and it's great to see so many developers here that are actually creating code at a conference that's being sponsored and put forward by Cisco. >> The AppDynamics is certainly a force in function for Cisco. Great acquisition. They paid a pretty penny for it, but it could've been more had they gone public (laughing) given the market that's in. So, AppDynamics comes in, Cisco now has a cloud-native direction, still work to do, but they're doing it in a way that's not all Cisco. They're not coming in with Cisco washing, "Hey, this is Cisco." They're really doing it right. They come in, 90% of the sessions are not Cisco at all. It's all community-based. Is this a working strategy for Cisco? Do you see, and what would you advise them because this is important. They have to do this, in my opinion. I think it's a great move, personally. But now, the innovator's dilemma is DevNet's exploding. You've got DevNet's beautiful community, it's growing. It's growing really fast. But, now you have Devnet Create. What do they do? >> Well, so, let's talk about the AppDynamics acquisition just for a second. As with everything, in talking about a big company acquiring a company of any size, you always have observe and see how it's going to play out, but it's got a lot of potential. One of the places that I think it's got a significant potential is in that AppDynamics, as a technology, does a great job of capturing metrics about application performance on networks and as we think about how the market and technology is going to be reconfigured so that networks can be better, more planful, more predictive about what kinds of things the applications are going to need, being able to surface that kind of data is going to be really, really crucial to setting the next round of conventions and that will lead to the answer of your question. If we think about where the market needs to go, we have full stack developers, we have networks. They talk to each other, but they don't engage in a meaningful way as often as they should and I think it's time for us to start thinking about above layer six, that layer seven, start breaking down layer seven and saying, "Well, that's where that full "distributed stack development's going to take place," so that we can start seeing how data will be reused, application services will be reused, componentry will be reused across a variety of different use cases and having that kind of a new structure defined and laid out so that it is built on the presumption that there's going to be a significant network in a way, in the middle of it, I think is going to be really important. Doing everything with RESTful APIs is really important, but I think the industry needs to get a lot more intelligent about how we're really going to build these things and not just presume that there is no network connection. There is a network connection. There has to be one and we have to build that into the architectures that we put forward in the future. >> So, programmable infrastructure as the DevOps ethos, that's what Cisco's proposing and saying they will and are becoming. I get that. I think that's the winning formula, but let's take that concept with what you just said. You're implying that okay, with now distributed infrastructure at scale, with AppDynamics and other things, the notion of a developer changes 'cause now, Cisco folks and their developer community, now is not just by itself, it's integrating in with the rest of the communities. That changes the notion of full stack developer because when you go hire, "I need to hire "a full stack developer," this stops really at the database or how low does it go? So, I think, you brought this up yesterday, not on camera, but after when we were kind of talking, is that this is an opportunity to reconfigure the new definition of a full stack developer. >> Yeah, I think so, John, and you know, one of the things we did talk about was when Susie was on camera yesterday, was the idea that yes, we can, as we introduce software-defined infrastructure, the infrastructure becomes programmable and so, we now see Cisco, CLI-type people thinking about programming instead of just doing command line work. We see it happening in the server world and the CAD world, et cetera, so there's no question that that notion of programmable infrastructure is becoming very real. What we're talking about is stepping it up and having it be available to developers in new and different ways, but utilizing new conventions that start to suggest that for time purposes, latency purposes, security purposes, think about organizing your application componentry in new ways so that the underlying network and infrastructure can provide even more robust capabilities and more consistent capabilities so that we can see further future ways of integrating these things together. And I think that's where this ends up. >> Interesting point about these network opportunity is that Cisco and networking guys are not a stranger to services. Network services have been around for a while. When you look at what came out of yesterday's conversation and this is consistent with a lot of our CUBE interviews we've done with cloud-native players like Amazon and everyone else like AWS and VMware and everyone else, is everything's a service. So the question that I have for the Cisco world is can they move quickly enough to a services model in this notion of a new network engineer, network developer, infrastructure developer? How well can they get transitioned over while preserving their core base of developers? >> I think they have to, so, and it's a great question and we're not going to have the answer from here, but I think one of the things we do need to start seeing and we're starting to hear rumblings of it is the idea that the network has to become more intelligent in the context of the services that the application developer utilizes that run above it and so, the network doesn't necessarily have to be made explicitly or overtly available, but it has to be intelligent enough so that it can provide new capabilities, new service levels, new security levels, et cetera, in a response to the way the services are invoked in the patterns of operation. You know, in many respects, we talked about this a little bit yesterday, John, I think we used to have infrastructure defined by hardware and that served the industry okay for a long time and in the last 10 years, we started talking about software-defined infrastructure. So, we moved from hardware-defined infrastructure to software-defined infrastructure and that's kind of where we are today with the idea of the network becoming more programmable. I think as a consequence of big data and recognition where digital business is going, where data really is the asset and the idea that we're going to build applications and then find data, we're going to start with data and then, decide what we need to do with that data through big data and other types of things, we're literally talking about, in the next five years, about something that we might call data-defined infrastructure where the data, the characteristics of the data, the location of the data, the way it's used, the way it creates value for the business, having a dominant impact on how the infrastructure gets configured and I think that's a, has enormous opportunity for Cisco. >> Yeah, Pat Gelsinger talks about the software-defined use and that's still part of the VMware strategy. This is kind of where it's going so I'm going to put you on the spot. >> Peter: Uh-oh. >> After yesterday's interviews, what did you learn? What did you walk away that's either net new information to you or something that validates something that you've been thinking about or had been researching and analyzing? >> Well, the first thing I'll note and somewhat self-servicing is Wikibon is known for being at the vanguard. We tend to be a little bit out in front and imagining what, how technology disruption's going to play out in response to the new use cases and business issues and it's always good to talk to people that are smarter than I am to start validating some of our positions. So, we heard a lot of, yesterday, that was pretty strong validation at a technical level and a couple of big vendors that are along the lines of what we're talking about so that was very useful. >> What did they validate? What specifically-- >> well, this notion of, for example, the notion of data-defined infrastructure. The idea that data in the future is going to be seminal to thinking about how infrastructure's intelligent, really configured based on the needs described within the data and the metadata. So, we heard that from a couple different people. Another thing that we heard was that there is a, that this not just, that Cisco's vision here of having developers and network jocks coming together to thinking about what the impact's going to be ultimately on how we create business value out of technology is something that's not just a Cisco pipe dream. We had four or five partners on yesterday, including a number of them who are quite sizable, RedHat, for example, who trumpet and reflect and are promoting similar types of concepts. And the other thing that I heard and I'm particularly going back to the PubNub conversation we had, it's really nice to see technology that is been around for a while, that works well, be really reconceived to be able to do new and different things and in particular, PubNub was talking about a deterministic Pub/Sub network infrastructure. Very interesting stuff, it's going to be really important. And the reason why I think that's important is because the lessons that we've learned in the past are not necessarily dead because we're going through transformation. One of the biggest things that I think we all need to take away from this is that we think about a computer, we think about my iPad, your Mac, a Dell machine, whatever else is, we think about that as computer, we think about a server as a computer, but the reality is, if you think about what we're trying to do through conferences like this is internet-scale computing where we look at the entire internet as a computer. Any data, any process-- >> Network's critical. The network is the essential element of it because that's what weaves the whole thing together and I think what we learned yesterday is the lessons of the past, some are going to be gone and we have to get rid of them, but a lot of them have more to do with business models. This is, we're still talking about computers and we're still talking about computer science. >> Great summary of your learnings. I learned, just to wrap up our intro segment, I learned a couple things. Observations that just popped out at me, one is Cisco has a lot of women in tech, engineers, so that was a very cool thing for me 'cause we always look at our index and theCUBE interviews over the thousands of people interviewed and still, only 18% have been women interviews. (laughing) Do more. That's our kind of passion. But they're smart and they're really knowledgeable and it's really awesome to see great women being featured. Certainly, Susie's a rockstar leader as CTO. The other thing that I learned in talking to Abby Kearns, Executive Director at Cloud Foundry and Dan at Cloud Native Foundation is open source is changing significantly-- >> Peter: It's a fair point. >> And open source communities used to be, "My community, yeah, we're winning!" And it's always been a gamer mentality or win-lose and I think now, with the horizontally-scalable cloud, you're starting to see a cross-pollination of players cross-pollinating and participating in multiple horizontal communities that together, is an expansion of the overall open source ecosystem. I think this is a new next generation dynamic that takes the tier one open source position, which really, it's our generation. We're seeing open source become tier one, not tier two, it's a tier one software where people's business models are now a open source. MuleSoft, these companies going public, multiple IPOs since RedHat just recently going public, I think you're going to see more business models on open source and open source is changing and I think cloud is a big part of that. >> That's a great point, John. And the only thing I'd add to it, since it's a great summary, is that we also heard yesterday that the very notion of leadership in the open source universe is starting to change. As people come up with new business models, they're also exploring with new ways of providing leadership that doesn't violate the basic precepts of open source because if there has been an issue in the open source universe, it's been that open source does a great job of, if you have a convention and a statement about where the product is, like an operating system, open source can do an equal or better job of it. If you start talking about use cases and a lot of business uncertainty and how open source can sometimes spend its time looking at each other, looking at other projects and filling holes in between projects and not getting to that use case. >> We've known for a while, it's not new to us that open source where the innovation is and that's certainly seeing companies have their employees there and that's where recruiting is going on, as well. But, I'm being more specific. I think the changing game in the open source community is going to be one that's going to reflect the structure of the industry and I think Cisco jumping in with this event will change the game in the makeup of open source and the projects because if you believe that the network is programmable, and that the cloud is one big computer operating system, then you have to believe that that's going to be a new domino that drops and falls and I think the impact of Cisco making programmable internet, programmable networks to developers will have a cascading effect that will ripple on and I think-- >> So, you're predicting a lot of new open source projects that Cisco's helped to catalyze. >> I'm predicting some turmoil that could be positive. Again, is it a collision or is it-- >> Peter: Yeah, it is. >> Edges coming together? >> And it's okay. >> John: And that's the chaos theory, you've talked about that. >> It's a good prediction. >> So, I think it's a lot of good stuff. We'll be watching and covering and of course, play-by-play action on theCUBE. More day two coverage on theCUBE after this short break. I'm John Furrier with Peter Burris, after this short break, stay with us. (upbeat music) >> Hi, I'm April Mitchell and I'm the Senior Director of Strategy & Planning for Cisco.