 Okay, we're back on the Cube. Day three, eight hours of coverage every day. We are here Sunday setting up. I actually flew out Monday night to go to San Francisco for the HBase conference and flew back this morning, Dave Vellante. We're everywhere, wherever the stories are, we want to cover them. Whether it's the tech alpha geeks in San Francisco talking about HBase or EMC, we're all talking about storage, big data cloud or talking provisioning with Puppet. Guys, welcome, Nigel Hurston from Puppet and Nick Weaver from EMC. Guys, thanks for coming on the Cube. Tell us about what's going on with Puppet and the deal with EMC because you got an apps dev ops like company playing in the elite, cutting bleeding edge cloud and EMC bare metal provisioning. So tell us the deal, how you got here and then we'll have a conversation around. So we just announced today that we've got Razor which is our next generation provisioning solution built by our good friends at EMC and we built it on top of some of the core technologies we have in the Puppet ecosystem. We've got automated inventory system, we've got automated node discovery and we've built this awesomely flexible system that lets you dynamically provision bare metal and honestly treat it with the same sort of agility that you have with the cloud and I'll let Nick talk more about how the project was developed. I mean, it came out of the fact that we were experimenting with bare metal live or with drill the ring study EMC and we kind of looked at a lot of different tools that were out there and we realized that not many of them actually partnered well if you're also trying to do dev ops after the bare metal provision, right? And there's a real kind of big gap and disconnect between the two. And so we said, you know, if we decided to do some of our self what would be the hit list? Like what would be the top things and how would you integrate it? And out of that, we decided to just give it a shot and work with it. And so we kind of packed up our little list of dreams traveled out to see Nigel and the guys and said, hey, what do you guys think about this? And the awesome thing about it was they basically came to us with this prototype and went, what do you think about this? And this is the software we would have built if we'd spent time working on it. So we just knew there was this really awesome synergy really quickly. That's nice. That's great. Like, did you flew by the Portland? Yes, I spent like five weeks, I think four or five weeks out there in the last couple months, yeah. Nice, great area there. So these guys came to you with software, great collaboration. You guys, hey, no problem. We can do this and just made it happen. Yep. How long ago was that? So I think the first time we actually kind of got permission to continue, it was about mid January. I think our first trip was in February. So between about in the February until now is about the total length of the project. That's like lightning for a big company like EMC, right? I mean, I mean. Yeah, it was very agile. We had a lot of help though. So we came in, we would come in with code and with designs and we'd balance it with these guys and they would, because they've got a great library of software. We were using their Factor product, which does the inventory. We got M-Collective, which is their distributed control framework built in. So they had a lot of great repos and knowledge and understanding. So we kept bouncing things off them and they kind of helped hone us and helped us figure out problems in short circuit issues. And so it was a great collaboration. A lot of design and influence from Puppet and then a lot of just coding and work from EMC. I think we really saw the benefits of having a really robust platform and this laser focus on just solving one problem really, really well. Exactly, yeah. Yeah, and I think, you know, EMC's got some work on, they've done with Pivotal, just done an acquisition site on the big data site. So, you know, from a coolness relevance factor, you know, it's really high. You guys could obviously have better with Opsco with Chef. So you guys are really doing well in the marketplace. This is the business model of the future for cloud, right? This is cloud. And it's on the hardware side, right? Bare metal. What's the environment look like with this solution? Is it just EMC solutions? Is it other servers? In fact, part of that list of dreams, one of the big line items was we should build something that can handle, you know, VMware virtualization, KMVM virtualization, any kind of physical hardware, any kind of blade, why corner ourselves with one product? It should be able to discover, inventory and understand and bring relevant metadata to make you make decisions with on anything. And that's why, you know, Puppet Labs factor was so important to us, is that's what it does, is it's able to inventory and extract all this great information, facts out of a system. That's why we use that. It's because we want to make it so we can have, so if EMC internally makes a choice on a different kind of platform to launch on, we're not restricted by choices we made earlier. So Nigel, when you talk to other vendors like Dell, I mean Dell's very active. I've seen OpenStack and a bunch of other initiatives in the community, but also they're trying to make their commodity hardware still be commodity hardware, but yet have some flexibility to integrate into these hyperscale environments like DevOps, Hadoop, and really cutting edge work where you can just stack, rack, and then expand and scale out. So that's great, anyone with Hadoop you can load a ton of stuff on one server, forget about the old way, but this is the new model. So commodity hardware is changing that definition. It's still kind of commodity hardware, but it's got a lot more software in it. So the question I have for you guys is, tell us your vision around how you see this provisioning, configuration management, automation side of the business that's been an app side integrating with some of the infrastructure hardware guys. Is it going to be running to APIs? Is it going to be, you know, how do you balance that arms race, if you will? Who goes first? Is it driven by the app guys? Are the hardware guys playing catch up? I think honestly, sysadmins and app guys are going to pick the software that works for them. And so I think that's why it was so important. This is vendor agnostic. It'll work with everything. And it's Apache licensed. There's people are free to innovate in an open source manner with their proprietary plugins. This space is as open as it can possibly be. And I think ultimately that's the software that wins. Talk about the openness because actually we're a big fan of open source. Explain to the folks out there who might not have a lot of depth in open source. When you put a project out in the open that could have agendas. Talk about what makes a project successful when it's in the open from the beginning. I think what makes it successful is fair governance. When people have an itch to scratch and they write code for it, that you're fair about accepting that code, that you're not blocking it because of commercial interests, but you're letting innovation proceed at its own pace. Is that balancing the commits, the committers, balancing that kind of committee of commit bits? What are we managing to commit code or is it? I think honestly, you judge on code quality and functionality. And if you're genuine about that, every successful product. Versus mailing it in with crappy code. No one wants that. It's meritocracy at this point, right? I mean, it's like, you know, good code speaks for itself, right? Yep. At the AMC standpoint, we went to Puppet Labs and said, look, we want to open source this because we realized that the best partner we could have in the DevOps community was Puppet and we wanted to adopt their model and the way they approached the community and the way they approached success, right? And so that partnership was, we went to them and said, look, how do you guys deliver functionality? How do you guys do open source? And they ended up coming back to us and said, listen, we do Apache license. We open it up to the community. We accept functionality. We accept features. And so EMC has made this investment but the EMC is going to reap the benefits also. So as people add functionality and extend this code. And that's the open source way. And that's like, we'll benefit from it strongly. Yeah, well, you guys, first of all, good citizen to doing that work. That's job one. Two, you're going to get the halo effect from the poll. Absolutely. And drafting off that success. The question Nick, I have for you now is, what were some of the challenges and complexities that you were looking to overcome with this partnership? When you started, you know, kind of kicking the tires and hey, I want to go out and talk to Puppet, what were some of the complexities you guys were trying to solve? I'd say it's really down to three things. One is already mentioned as a big air gap, right? So everybody looks at bare metal as this single task. It's a long operational cycle. It's painful to do. It's kind of like, if you're an application guy or administrator guy, sometimes it's one of those things where it goes into a black hole, right? You say, hey, I need you to redo the server. I need 10 more servers. Goes off in a black hole. Nobody knows what happens to it. Then all of a sudden two weeks later you get your servers, right? And then the second piece we had was there was this real big disconnects with tools that do provisioning and then the upper level stack control. So Puppet has the ability to deploy with our demos vSphere, right? It could be not right now, but also OpenStack, Apache, Hadoop, all these great platforms on top, but it has no visibility down into the bare metal if the actual thing's underneath that, right? So the issue becomes if there's some way we can actually harvest the information at bare metal and share that while provisioning with the upper level configuration, you can make some really cool decisions, right? So you can take finite hardware of different sizes and shapes and Puppet can take it and go, okay, this is how I want to slice it up into a proper stack, right? You know, proxy servers are a little smaller than the big object servers. You can take mapping and do all kinds of cool stuff with these very complex application steps to can inherit finite resources. So what's the next step in the evolutionist project? Is it going to be an open source project? Is it going to be managed by Puppet? Is it going to be, how are you guys looking at this? We're combining governance on the RAZR project and it's going to continue to be open source. We are a really strong believer in the fact that the components of the platform for your infrastructure really have to be open source. There's lots of scope for people to have commercial differentiators on top, but really the plumbing itself needs to be open source. So this is all going to be headlined under RAZR as the project, okay, great. EBC actually took the code and gave it over to Puppet Labs to go underneath their license, so it's properly underneath all the rest of the Puppet. Okay, cool. So you guys kind of be the gatekeeper, that's cool. So Nigel, give us a quick update on what's going on Puppet with the company, some recent status updates. You guys were on VMworld, we had you on. Sure. The company talking about what's going on in your product and company. Sure, so we've been continuing to release Puppet Enterprise, which is our commercial version where we take the well-tried, well-tested open-stack components, open-source components, build out the stack and provide a bunch of commercial differentiators on there. We've got some great GUIs around orchestration, live management, which allows you to visually browse all of the resources on your nodes and then clone them across machines. So people who don't even have any idea how to write manifest, how to write any code can really leverage the power of the platform. Awesome, well congratulations, Nick. What's next for you guys? You're going to continue this relationship, obviously, with EMC. Does this mean you're going to get an apartment in Portland or travel out there? They have good beer in Portland. They keep offering you guys. They say I can office out of there if I wanted to. The best beer is up in Portland. You know that. They do coffee too, actually. Coffee, beer. I'm being impressed by the coffee as well. That's why everyone's so chilled out up here. They drink a lot, then they have coffee in the morning. And the food carts. I'm a huge fan of the food carts, actually. A lot of innovation in Portland. Alex Williams, our contributor's up there. Clint Finley, a writer for us, was up there as well, so good stuff. So the next thing, I mean, we still have a lot of stuff we wanted to get into the code. We haven't finished yet, so look for EMC to add some more features and more value. And I don't want to say too much ahead of time, but we have a lot of ambition with the code. All of it going into the community, all of it benefiting everybody. But this is just a tip to the spear for us. All right, guys, thanks so much. We're going to be right back to wrap up EMC World Day Three. We have waiting for Scott McNeely to come on theCUBE, and that'll be a fun interview for David and I as old school dudes. We'll have a good wrap with Scott McNeely to talk about the Sundays and the Unix. So we'll be right back after this break.