 from the Computer History Museum in the heart of Silicon Valley. It's theCUBE, covering OpenStack Silicon Valley 2016. Brought to you by Morantis. Now, here are your hosts, John Furrier and Lisa Martin. Okay, welcome back everyone. We are live at Silicon Valley for OpenStack SV. This is theCUBE, Silicon Angles flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier, my co-host, Lisa Martin. Next guest is Christian Carrasco, cloud advisor for cloudguide.io, advisor, CTO, serial CTO, and here at the OpenStack SV, giving the keynote on why OpenStack is failing controversial topic when everyone thinks it's thriving. Welcome to theCUBE. Thank you. Thank you for having me. We'd love to controversy, because one, you know, the glass is half full, half empty, but certainly there's a lot of believers here. Yeah. And so you have that group thing going on. Sometimes it could be toxic or just good motivation. I actually like to think they're doing well, but you got a perspective from the trenches. Share your thoughts on why is OpenStack failing? So, basically, I think it's actually succeeding, but there's a lot of failures out there that are highly publicized, right? And so we are currently, I'm advising the company in building, re-architecting their cloud, and they have about half a billion users in this cloud. And in order to go to the next level with their cloud, they had to look at OpenStack and definitely the public providers, and we looked at the whole spectrum. And the research that I was doing, there's a lot of failures in OpenStack. People abandoned it, they tried it out, and they left. Right? And so at the keynote that I gave, it was basically summarizing the main points, not all of them, right? It's not going to be concise ever, but the key areas that I'm seeing that are big in the industry, why it fails, right? And so, and the main part is, it's not the technology most of the time, it's going to be everything else based on human intervention. All right, so I'll take this through some of these, so you agree it's healthy now, somewhat stabilized, OpenStack as a community, but the failures are of the deployments. Right. How far back are we going time-wise? So, recently failures, going back a year? It's actually over the past five years, right? So, an example that I talked about is Dogma, right? It's, people develop a view of a technology, they go and they try it out, it's version 0.5, and they say, this is terrible, we're never going to use this again, because they had a bad experience. That's what happened in OpenStack, very early on. It got a lot of traction, engineers went and tried it, and they hated it, and they moved on, right? And we still see that. It's a bad taste in their mouth, and they make a decision. Exactly. What I call dogmatic tech views, right? We develop them, and then we never touch it again, because we got burned, right? Other reasons, 3AM, PTSD. And my previous couple startups that I've had in Exit, where I had PTSD at 3AM, because I would get woken up every other night, because of this, and that's what actually drove me into the cloud, and or high availability when it was called or distributed computing. So, your recovery must go to the cloud. Exactly. That was your T-talks. That was my therapy, you know? Yeah. And so, that's why I've been very close with the cloud industry for a long time, and I know it can succeed, but it's just got these issues in there. I think, you know, we look at failure as one of the efforts, but I like your spin on, you bring advice, basically, distilling that down what you talked about this morning. I'd love to get some of that advice for not, as you say, it's not a technology problem. It's more dogma, it's PTSD. It's also bad experiences. What is your advice to clients, TapDoy may be included, in terms of how they particularly hire talent that doesn't come in with those biases, those bad experiences. Do you recommend that they hire talent from different industries, or talent that's younger, fresher? What's your recommendation there? Right. I think the cloud operations department, which is really, you have to start with a head of your cloud department. Right. Today those are called either cloud officers, captain of the cloud, chief cloud division manager, whatever you want to call them, there's tons of names, but you need a known, you need the captain to direct the show. From there, that captain will look at your business model and make sure that your cloud initiatives align with your business initiatives. That may mean that you have to, for example, in Dropbox's case, go design hardware for your storage, or work with a hardware partner. So they brought in hardware people. Other times you may need industry leaders in the cloud that are great engineers who are writing code for OpenStack because you need customization. And then for other companies, it may just be people who understand how to scale between public and private simultaneously. And so we're seeing that. So it's not, there's no one answer, what's the right formula, but the one key ingredient is it's not a side project that's owned by your DevOps team, by your IT operations team, or some guy in the company. It's a head captain of your cloud initiative who will then decide what you need to do. Is that a common problem with those enterprises who have failed? Are you seeing that it's been simply a lack of that single point, that cloud captain, that driver? I would say a lot of those failed initiatives were because of not having the right captain or not having the right captain with experience in the industry. And talk to us about what you've done with Tapjoy. You've talked about re-architecting that cloud. Was that one of the first places that you started to put that leader in place? Or what were some of the complexities that you've helped them to overcome? Yeah, so the issue was there was obviously, there was actually, they've made a big investment in OpenStack very early on about two or three years ago. And so they're moving in that direction, but they didn't fully finish moving in that direction. And now they're explaining, okay, what do we do? What's the next step? That's where I came in and they actually realized, we do need some help in this. This is not our area of expertise. Let's bring somebody who can advise us and help us build that and put in place. So there is, the architecture includes both public, private, and having the right team. We're looking at hardware stuff. We're working with several people from here and working with managed services. In TapJoy's case, they're not really going to build out an entire engineering team. They're actually going to outsource some of that. And so it's going to be a mixture. That's what they found was the best fit for them. And there's tons of companies that can help out and fill in all those gaps. So I was going to ask you about some of your philosophies on cloud. Obviously you've got a good view on it. You've been there almost as a way to kind of, hey, I'm done. I got a scar tissue on the on-prem stuff, but cloud obviously has a lot of benefits. But it's growing. You coined a term, maybe you did or not coined a term, hyperconverged cloud. Certainly we love hyperconverged. We've been covering hyperconverged storage, networks, our Wikibon team. What is hyperconverged cloud? Can you explain the concept and where does it fit in for the future? Yeah, so I think, and I purposely grabbed the hyperconverged, the first half of that, and put the cloud in it. The message that I want to portray is you've got to get a little bit out of the weeds. We've got to go a little bit higher. Let's look at the entire cloud picture. All, you know, virtualization, Docker, OpenStack, Googles, the Amazons, everybody. Let's get together and converge it completely. And when we converge it completely, we're talking about your cloud, not their cloud. We own the IP, right? And then we can power that cloud with any provider we desire. Right? So create a centralized Uber cloud, for lack of, not we'll say Uber, but a mega cloud for the customer owns, using a power source, if you will, compute source from whatever provider that might, like Amazon could be pitching in connected to it. Yeah, more from the IP perspective. It's not really, there's no compute within it. So I'm looking at your cloud as almost like a brand, right? Your cloud, what happens right now, you go to a public cloud provider, and you put in all their services and you're tied into their, and you cite your software application code, is their DNA, right? You can't pull your cloud out of there. So is that really your cloud? If it was your cloud, you'd be able to pull it out. It's Roche Hotel, you can check in, but you can't check out. Exactly. And so, so the goal is, let's get everybody together, right? And create standards, interoperability standards. And the reason I believe that's a success to the next generation cloud is because we go back in history and we look at other successful technologies. And for example, the PC industry, what happened with the PC when they standardized interoperability between competitors? Well, it exploded and it became a household device. So this concept really puts the customer's IP as the driver, not the follower to somebody else. Correct, yeah. So they forced the interoperability, almost reversing the psychology on the provider, saying, if you want my business, you have to interoperate. Exactly, right? And it creates a better market, more competition, better services. What I believe it's happening to in the cloud industry, they're niche providers. So they're not competing with the Amazons and the Googles or the open stacks. They're saying, we'll plug in this particular service that we can do really well in our cloud with our technology and plug it into your cloud and you can enhance it, right? So who's doing this right now? Do you have any examples? I mean, we interviewed on theCUBE, a company that has a global cloud contract. Right, well, what I, so yeah, there's the one that I just, I looked at and I thought it was a pretty cool idea. It's a machine zone. They do gaming, they're a gaming company or I think they're relabeling, they're rebranding themselves and they launched a, what they originally called a cloud service using their real-time engines, right? Like the real-time computing, I'm probably going to butcher what their real definition is but it's basically real-time computing, real-time engines for their gaming platform. That is an awesome service that you could use for other gaming devices, right? Other gaming companies could really bring that in and put it inside their cloud or their devices, whatever that may be. However, I think they're thinking even bigger than that, right, and they're looking at other industries that can benefit from real-time computing. So real-time computing is very difficult to do and they've somehow mastered it and figured out a way to monetize it. So they're putting that, so they're not competing against Amazon, they're not competing against OpenStack, they're not competing against the traditional cloud, public and private clouds, they're just enhancing that, right? And monetization is actually one of the things that we talked about yesterday, John, is where OpenStack and the open source community is now. We talked with some guests yesterday who are figuring out new monetization streams. But I'm curious in terms of business impact, can you expand a little bit more on re-architecting TapJoy's father, a mobile app company? What's the business impact, the business diet that TapJoy is getting from under the hood re-architecture? What's that impact that they're making? Actually, there's several impacts to TapJoy that a re-architecture like this would benefit them. A lot of times people dismiss the fact that it's OpenStack is actually less expensive. That's a huge benefit. I mean, we dismiss it because engineers don't really care about the financial costs, but at the end of the day, the company's in business to make money, right? And so that's a huge impact, right? But we also have to look at other areas, like engineering, right? If you own your cloud, the flexibility to do other technologies and customized technologies, the fact that OpenStack is customizable, it opens up endless opportunities, right? And they have the engineering resources to modify and create whatever they want. So that opens up a slew of possibilities. And then finally, we look at, again, machine zone as an example, and we look at what they're doing and TapJoy could also position themselves to grab a particular piece of technology that they're good at eventually and also start offering a service, right? And become in there. And that's obviously- I mean, the microservices is just an opportunity for them not to have to build software if they can just get it from somebody else, either from APIs or whatever. So, yeah, correct. I mean, there's tons of opportunities. So I mean, the sky's the limit really. And again, it's not about just the private versus public, we will need both always. All companies will always need both. It's about positioning yourself to have the flexibility to turn and respond to the market demand. Right, and be able to reduce costs, to be able to focus on customer acquisition, customer satisfaction, and gaining share. Exactly, yeah. Christian, thanks so much for coming on theCUBE and congratulating us on a great talk. I want to give him a plug up because he's got a new Twitter handle. He only has seven followers. So, if you're watching this follow Christian on Twitter, it's the Twitter handle CloudGuy underscore IO. The CloudGuy IO and the great concepts. Love the hyper-converged topic. I think that's a great trend. I think you're right on the money on that one. I think that is a good future. It's essentially a data center in the cloud with new services, kind of SLAs can be built into it. And the bigger players, certainly, I think would go to that model. I mean, it makes a lot of, basically it's hyper-cloud. Yeah, exactly. It's hyper-cloud is a product that we can't say is a product yet. So, awesome. Congratulations. Christian, thanks for coming on theCUBE. Appreciate it. This is theCUBE live in Silicon Valley. I'm John Furrier with Lisa Martin. This is theCUBE. You're watching live in Silicon Valley. We're right back with more after this short break.