 Okay, we're back here live inside theCUBE. This is SiliconANGLE and Wikibon's theCUBE, our flagship program. We go out to the events and extract the signal from the noise and this segment is day two and we're going to break down with our chief analyst, David Flawyer, who's on the ground here inside theCUBE to share with us what the hell's happening in the sessions because there's so much signal. Not a lot of noise and we can't keep track of it. We need another, we need another cube. We need another analyst, Dave Vellante, if you're out there, you know, come out, come out here. I know he's watching. Shout out to Dave Vellante. He's tweeting. My co-host is because Jeff Frick, Jeff, David, let's break this down real quick. First of all, horses on the track, okay? Let's get the horses on the track and then we'll dive into the sessions. So, David Flawyer, first to get here from you. Who are the main horses here that are running in this open space? I see Rackspace, they were on theCUBE. I see a big proponent of it and props to Rackspace. Big halo effect, no need to drill down there. What are the horses we have on the track? So there's a combination of categories. So first of all, there's the vendor category itself. So you've got key players like IBM, like Dell, like EMC in conjunction with Cisco and Cisco themselves. You've got those, as you've got HP with a recent moonshot and that converged infrastructure. All of these are offering converged infrastructure in one way or another, either as reference architectures and or actual products themselves. You've got VCE from EMC and Cisco. You've got Pure from IBM, et cetera. You've got all of these players in this space and they are providing economies of scale. They are providing a reference architecture for specific workloads, like for example, an SAP installation or a VDI installation. So they're offering economies of reduction in the complexity and the speed to deliver. That's one thrust that's going through there. The other thrust is people like Google, people like Amazon, people like Yahoo, people like Microsoft, Azure, who have a huge amount of compute power and are providing access to that compute power and those services and are delivering other services, a lot of them consumer to begin with, but increasing the enterprise services, if Google, for example, providing all of the services like Google Docs and all of those services the same way as Microsoft is. So there's another set of services coming from those sets of people. And then the third are the open, like a rack space, are the open compute foundation, the rack space foundation, eucalyptus, and a few others as well, who are providing a cloud architecture, a cloud platform on which to develop these sorts of applications. And of course, VMware is very much in that space as well with there. So I think those are three categories which is useful to break down because there's a lot of similarities. And what was the second category again? The first was users, the second was the traditional cloud vendors like Microsoft, Google, and the others. And then the cloud vendors in the middle of the day. The users compute power on task. The guys have computing power. But she said something very interesting, David, the last segment, as you talked about separating the hardware from the software before you had an integrated unit that solved the problem and now you're separating it. But the flip side of that coin has always been if you try to run best of breathe then you always have an extra layer of integration. And there's a whole different kind of management thing. So what's happening now that's different that's kind of swinging the pendulum back towards separation and best of breathe? Well, the key is that what you want to be able to do, the model originally was, I'm the developer, I decide what is going to be developed on. I will design the hardware and my development software to run as a unit. And that was a vertical stack if you like. That was the application, either the ISV or the developer, in-house developer. That was the way that they did things. And what's happening now is the DevOps is much more about how can I separate out the software, make the software run itself, the application itself, and then point it at a hardware stack which is a set of resources, compute resources, network resources, storage resources, and separate out those two. So you've got a hardware stack with those resources, you've got the application which is managing itself and assuming that it's going to have to go and find those resources. So it's a different model. Now that model is what cloud providers are using. That's what all modern platforms are starting to move towards very, very quickly indeed. And the advantage of that model is that it is self-tuning, it's self-providing. And the other advantage of it is that it's much easier to put into the cloud and much faster in its turnover. You don't have to worry about the hardware side of it. You can do something every three months, every six months, as opposed to the old way of doing things with ISVs, but at the time they got through the whole of the stack, it's a two-year cycle. So the rate of change and the rate of adaption is much, much faster in that environment. David, I want to ask you a question. You and I were talking prior to the morning when we kicked off that we were both amazed, but you made a comment I want to share with the crowd. This is a vendor hype-free zone and there's a lot of real users stepping up, demonstrating, sharing, proud and loud what they're doing with OpenStack. Can you elaborate on the comments and then talk about some of the use cases? And this is not like, you know, small and never heard of this.com company, but like big names, so talk about that. Yeah, I mean, what we had the privilege of with the Analyst Day was one of the best Analyst Days I've ever been to, if not the best. And it was a very simple introduction. Here's OpenStack, here's the principles. And then here are the people using it. And so they were able to go straight into the meat and potatoes rather than have this highly orchestrated vendor type of communication. So it was an excellent day. So to give a few examples, for example, Best Buy, they are one of the largest retailers in the world. They have a huge peak in a Thanksgiving day of every year, seven times the average. I mean, that's a huge peak to have in the middle of the year. So they needed something elastic, something that they could expand into. They needed to be able to build architectures which, A, on one side, were much more intuitive for the end users, gave a lot more help to the end users. And secondly, could expand to meet these huge volumes that they were getting. So that was one example, and they've put their new work onto OpenStack, a new roll out of some of their services. One of the key metrics there is that traditionally it takes them 30 seconds or so to get out a page because it has so many additional services that it's added onto. They've taken all of that, brought it in-house, brought it inside the cloud, they can now render that page in two and a half seconds. So from an end user perspective, it's now like a Google page. It comes up straight away within seconds as opposed to being something that is, comes up 30 seconds later. So let's break that down. So in our world, we talk to all these big whales, IBM, Cisco, HP, Dell, EMC, NetApp, all the top companies. And they have an old architecture of all. They come from the old generation and are actually part of inventing the news. So those guys are doing a good job. But that world that you just described with Best Buy was hard to do five years ago. Databases were different. Converged infrastructure is now software-led infrastructure. Can you explain some of the under the hood things that are relevant in the hot areas that are the most explosive from an innovation standpoint? Well, the innovation is happening in the cloud. These new ways of designing applications. One of the most important innovations now is much lower latency on IO. So instead of having the old fashioned IO which took milliseconds to do, you can now, or tens of milliseconds to do, you can now get down to IO's of one millisecond much lower than that. Actually down to now to the microseconds and even into the nanosecond space. Now that has a huge implication on the quality of the application that you can make. If you've got all of this compute power and you've got all of this data, now suddenly you can make applications which are much richer in the quality and do more. And you can make your database structures much simpler, much flatter. And instead of having to run lots of different modules and having to the workflow dictated by the flow of the modules, you can take a flatter database and you can define your workload onto that database. So I want to ask you because there's a company that's in Silicon Valley, NetApp, who's in that you follow. They have a cool expression called Agile Infrastructure. Yeah. Agile, I think it was called Agile Data, whatever, Agile Infrastructure. But that is a core thesis of theirs. Agility, Agile Enterprises. At the same time, they also have a phrase called non-disruptive operations, which kind of doesn't mean anything. It's just more of a management consulting phrase, but it does, it's categorically mean something. Could you talk about non-disruptive operations from the sense of what you just talked about? I mean, Best Buy has to have variability, agile capabilities to serve and power their business model. At the same time, there's still the emphasis of non-disruptive operations means we've got to have availability. So could you break down what non-disruptive operations means from an industry standpoint? Well, so you've got planned outages and you've got unplanned outages. Well, first of all, you have to build the application assuming that there are unplanned outages. Oh, for the best plan, for the worst. Well, no, a server is going to go down. Your storage is going to go down. You have to build that into the structure of saying this service is not 100%. This is what I do. As a design criteria. As a design criteria of the application and the infrastructure within it. So that's the number of the first thing you have to do is assume that it's going to go down and have sets of extra compute and data resources and network resources available to you. The other point of the architecture is that you've got to design in the ability to have planned outage. In other words, that you have rolling changes. So beforehand, it was a change every two years. Now, it's a change every three months. And that's again a major change in the design that you have to apply to it. Which means that you have to be able to swap out the hardware. So non-stop outage doesn't mean high availability. It means basically plan and have an architecture. Assume that you're going to have to change on the fly. Whether it's software upgrades or other things. Absolutely. Everything has to be applied in that way. Tell me about what is the definition of in your mind agile infrastructure? What is the, what is, what needs to happen for an infrastructure to be agile? Support developers, it has kind of a developer angle to it. I mean, what does that mean? So the developers and the analysts themselves have to be working together to supply an ever-changing, quick, rapid ability to put in new levels of software, new levels of hardware, and ensure a seamless transition from one cloud to another cloud to one version to another version. As David Vellante always says, you're our Peter Gammons of the tech business. You come in and give the analysis on the inside baseballs. But I have the final question before we break is I want to ask you is, in your expert opinion and surveying the field here and looking at the horses on the track, so to speak, and knowing what's happening in OpenStack, AWS and stuff we're covering in software infrastructure, how much impact will DevOps and infrastructure as code have on existing legacy are all environments right now in the industry? And the data center in particular and service providers. So the impact on current applications with the legacy code within that, it'll have very little impact at all. It's going to be for new code that's coming out and new applications. So the interesting thing is what do you do about those legacy applications and how do you bridge that gap between today's environment, mainly legacy, to tomorrow's environment, mainly cloud-based services? So one of the ways, there are many ways of that can happen. The hybrid cloud is one of those architectures that may contribute. But personally, I believe the better way of doing it is to move the traditional IT resources into a co-location facility, into a mega data center, outsource them and have the cloud services that you're going to use and move over to in that same data center so that you minimize the data flow between them and then maximize the speed at which you can migrate applications to that new environment. And I'm going to put you on the spot for a final word here. Of all the big whales we've been talking about, who's best positioned for this new world? Well, clearly Amazon has got a lot of meat behind it. Clearly VMware have a lot of marketplace to lose in this space. But people like Rackspace a year ago, they would have been still way at the back of the pack, they have accelerated and come forward at a huge rate. And how about the Cisco's EMC's net asset of the world? Where are they standing, IBM? Well, they are going to win anyway for a long time with the traditional IT. These things don't happen in one year or two years. These are five, 10 year horizons. Still they do very well. They have to learn how to work in an open environment, how to contribute to it, and they are very fast. And they're here, they're putting a stake in for it. Exactly, they're here, EMC is here, and IBM is here, NetApp is here. They're contributing to this. Okay, David Floyer, again, disruptive technology, disruptive transformative environment, new brands are emerging, new names, startups, but also the big guys are here. They're not going to lose their market share easily. They're going to reinvent the new and be part of it. So we'll be right back with one of those in a minute, we're going to hear from HP inside the cube after the short break. We'll be right back to SiliconANGLE's exclusive coverage of OpenStack Summit in Portland, Oregon.