 Live from Las Vegas, it's theCUBE. Covering Edge 2016, brought to you by IBM. Now, here are your hosts, Dave Vellante and Stu Miniman. We're back, this is theCUBE. This is, we're wrapping up day two here at IBM Edge. This is the fifth year that theCUBE has been to IBM Edge. theCUBE is the worldwide leader in live tech coverage. Major Hayden this year is a principal architect at Rackspace. Good to see you, thanks for coming on. And thanks for having me. So you build OpenStack Clouds? A few times, yeah. All right, how's that going? Very well. So I work on our private cloud product at Rackspace. We call it Rackspace Private Cloud, pretty good name. And so what we do is, all over on the joke. So what we do is we go and build private clouds for customers to their spec, whatever they need. If they need storage change up, we change it. So a lot of what I work on is the open source deployment components with the OpenStack Ansible project within OpenStack. And what it does is takes all the components, takes Nova, Glance, Sender, everything, and deploys it to spec. So if a customer gives us a spec and says, I want Sender configured in this way or I want to configure it with the storage device, we can go and configure that and do that deployment form and then maintain it going forward. So that's just plain Ansible. It's an automated workflow deployment system, right? So if you're familiar with Puppet or Chef or Salt or any of the other ones, Ansible does a lot of the same things. So it'll deploy configuration files, it'll start services, stop services, and do quite a few other things in selling packages. But I think the reason that we really like it is that it's Python based, so it's OpenStack, so it makes it very easy to go back and forth. The other nice thing is you don't have to worry about an agent on the other machine. You know, you just actually just run the deployment as long as SSH is running, it works. You don't have to worry about if there's an agent on the other side that is up or down or not behaving or anything like that. All right, so Ansible was acquired by Red Hat, so no concern there. Obviously Red Hat's going to keep it open source. That's their whole mantra. I certainly hope so. No, absolutely. Jim Wider says, never again will I do anything that even smells like it might not be open source. So, yeah. No, we love it because it's something we can build upon too. So if we find something in Ansible that's very close to what we need, but we just need a little something extra, we can go and write a plugin or we can write a module and just add that in and deliver something for a customer a lot faster. And you've personally made a lot of contributions to open source, right? What's the state of OpenStack today? How would you describe it? Update our audience. So it's growing like crazy. Every time I turn around, there's a new project that's doing something or we'll be sitting there at the office and say, I wish we had something that did this and we'd look around and be like, hey, there's one that started six months ago, maybe we should go take a look. So for me, I feel like it's really vibrant. Like there's a lot of the small stuff that we used to sweat that we don't sweat anymore. Like we have agreements upon requirements. We have agreements upon how things are configured. Configuration options still are a little bit crazy, but a lot of this low hanging fruit that we used to deal with is now gone. And so we can work at lower levels within the stack a lot faster. So if you need to add a plugin to Neutron or you need to have Nova do something that it didn't do before, there's already kind of an agreement on how you should do that. You don't have to kind of step around and figure out how to do it. I guess, can you help us kind of unpack that a little bit, because when I talk to the community, simple is not the term word that comes up when people talk about open stack. It's not something that people are just going to go do it themselves, but a lot of ways that's probably where Rackspace fills in some of the gaps of, where does Rackspace fill in the gaps today? I say, where is it things that you put it together and where is it code? Where is it deployment? Where is it knowledge and fanatical support and all that? Right, so open stack is made of a bunch of parts, obviously. So what we like to equate it to is, you wouldn't go order all the parts to make a car and make it yourself. You're going to go down to the car dealership, you're going to pick one that's available that you trust, from a vendor you trust. And so that's what Rackspace does. So we spend our time in the community understanding what's going on, listing the people and the projects, and then what we do is go and make really good choices on how we deploy everything. And then from there, if there's things that change or if a customer has new demands, we try to make really good choices around how that should be deployed. So I guess the long and short of it is, is that if you have a vendor that knows the open stack ecosystem well and knows how to put the parts together, it can actually be pretty simple to operate. But if you're just coming at it from the get-go and trying to grab the parts and stick them together and figure out how to get the hardware and the hypervisors together, it can be kind of a rough experience. Yeah. Where is the solutions that you're putting together helping customers beyond just the infrastructure layer? I think the hot topics we've talked about at this show, things like security, networking tends, networking and storage both tend to be kind of boat anchors for a lot of IT infrastructures. How do the Rackspace solutions help there? So first off on security, within our product, and we've done this in open source, is that we apply security hardening for you in an automated fashion. So it runs, takes about a minute to 90 seconds to run during the deployment, and you can fix it and forget it. So when your auditor shows up and says, hey, it's time for your PCI audit, are you hardening your systems? You can show them the document, we're hardening to the spec, here's our exceptions. And I think the rest of it around networking and storage is being able to, instead of going to a customer and saying, what are the myriad of options that you want to choose from? Maybe we have a small set that we recommend and say, okay, you can have storage locally on the hypervisor, or maybe you hook it up to a sand device, or maybe you have Ceph, or maybe you have sender, just like a plain LVM type sender. So we go to the customer with scenarios, and of course if they need to customize from there, we can go and do customizations. But we can say, hey look, based on what you told us you want to accomplish, these two scenarios might work for you, and then from there we can make small customizations, instead of going fully custom from the start. Yeah, so we're quite bullish on open source, love to support all the solutions out there. We've definitely heard from customers gone from the, oh, there's open source there, to oh, I want open source. How many of the clients that you're working with are not only asking for open source, but actively contributing to various projects? Quite a few actually, and it's growing, slowly but surely, and sometimes I'll find, for example, in our OpenStack project, OpenStack Ansible that does all the orchestration. We'll have someone show up and say, hey, I'd love to contribute code, and we look at where they're from, and they're like, oh, this company's huge. You know, do you guys use OpenStack? And they're like, yeah, we do, and we found your stuff, and we want to start using that. So it's really exciting, and I think the thing I like about that the most is that we find out these use cases that we didn't think about before. So we might say, oh, look, this will never be a bottleneck, and someone will show up from, I don't know, cable or a telecom vendor and say, oh, yes, that is a bottleneck, and then we can regroup around that and try and figure it out. So you're talking about a big financial services institution? Potentially, yeah. I mean, it's true, you go around and give these talks and say who's using OpenStack, and you're surprised sometimes at what they're doing. Definitely, yeah, so there's a lot of people who do Dev and test, like for example, like maybe you have an IT department that doesn't want to go and rack servers constantly. They can go rack a bunch, deploy OpenStack on it, and then allocate that out to people. So if marketing needs to test drive a marketing campaign, they can do it. If finance needs to go and crunch a bunch of numbers, they can do it. And then you can kind of do that back billing or some people like to call it shame billing, you know, where you go to them and say, look how much you've used. But at least it gets you out of the data center. You're not racking a box, everything's software defined. All right, so let's talk about Rackspace, the company for a little bit. There's been a lot of changes over the last few years. You know, it's a certain piece of the business that's sold off, big news going private equity, which is, you know, big mega trend that we've been watching everybody from, you know, Dell and Riverbed and, you know, many others and Rackspace, talk to us a little bit. You've been there for a bit. You know, what's going on there? What's the kind of the big, you know, overarching strategy? Yeah, so I mean, I've been there for 10 years and we've gone through lots of changes in 10 years. So I've been there when we were private and public. And so one thing I'll say is what we always try and do and one thing that we do is we always reorganize ourselves around what customers need. And of course that changes a lot, you know, as technologies show up or as different demands from customers show up. And so I think one thing we've always done is really try to figure out like, what do the customers need and how can we be there to serve them in the best way possible? So some of the newest stuff we've been working on is saying, we'll support you wherever you go. So if a customer says, hey, I love this offering from Amazon. I want to go there. We'll say, hey, we can help you with that offering. We can help you manage it. And then we can figure out how to deploy it. So a good example I like to give is if a customer says, hey, I need a database on Amazon. Well, we can go and give them a recommendation of putting the database server in EC2 or using the minus QL as a service and then kind of the trade-offs you get between each and the pricing. So I think really, I guess the summary of that is we try to adapt to what our customers need. So if there's changes we need to make, if there's, if we need to shift things in the business, we make those shifts for the customers. And what about, what's the relationship with IBM? What are you guys doing with those guys? Is it a power play? Yeah, yeah, so open power. That's true, yeah. So it's been talked about a lot at the conference already but we're very involved with open power. I think one of the things we've seen, I think it's been beaten to death at this conference but Moore's law is kind of going away a bit. I guess that's an understatement. I saw some tweets. Quite a few. We're going to say it one more time. But yeah, I mean all day yesterday that was the thing, Moore's law is going away. So you have to have more specialized hardware for specialized workloads. And so we at Rackspace have looked into power as one of those things where combining power with CAPI, the accelerators, being able to give customers greater performance, the performance that they're looking for that wouldn't necessarily come from a chip refresh. So if a customer has a database workload or some type of computational workload, we can find a way to accelerate that and say instead of waiting for that next processor generation, let's accelerate that workload to where you don't really have to worry about it anymore and you can go focus on something else. And so open power is something we've been involved with and we created the barrel-eye board which I believe you can actually go out and purchase from a vendor today. And we're consistently working on that with IBM. Great. All right, well thank you very much for coming on theCUBE. We'll give you the last word. Final thoughts on what's next for you guys? What's next for you personally? Where are you headed? Well, I definitely think there's more on power. I'm excited about these specialized workloads, like finding the exact machine that can get you that performance that you're looking for and taking those things that almost you're like you were afraid of, like I want to do deep learning but I don't have the computational power for that. Well, what if you did? And really trying to figure out how to get that to work. Great, excellent. Well, thanks very much for coming on theCUBE. Major, I appreciate it. Thanks. All right, keep it right there, everybody. Stu and I would be back at the wrap day two of Edge right after this is theCUBE. We're live from IBM Edge, right back.