 All right, I think we're ready to go. So thank you all. Thank you all for coming. I can't see you. These lights are glaring at me, so. So I think we're giving folks a couple of minutes to come in. So I appreciate you being here. Hope you're having a good summit, a great summit. My name is Ashish Nardkirni, and I'm joined by my colleague, Larry Carvalho, where part of IDC's research team. I'm part of the infrastructure team, and Larry is part of the software team, and he covers platform technologies. And I'll let him provide more details on his research area. But my research area is primarily infrastructure-related storage, computing, and software-defined infrastructure. So we are here to kind of talk about IDC's research related to OpenStack, and it's part of an ongoing umbrella research, which we call a software-defined infrastructure. And we're here to sort of talk about some of the findings at a broader level, why people are looking at software-defined, and what are some of the key benefits that they try to obtain from going into an infrastructure that is software-defined in nature. I'm getting used to this, so please bear with me. So yeah, so we have literally split our presentation into two sections. There's the software-defined infrastructure piece, which I'll cover, and then Larry will cover Cloud and Open Source. So he has some research that is kind of related to that. So just to sort of set the stage, we are talking about an evolving data landscape, and certainly on stage today, we talked about the different types of applications, different types of applications creating different types of data sets. And it's all about moving the data and compute closer and closer, so the movement is minimized. We need to kind of run the workloads right where the data is located. We want to have a geodispersed infrastructure, location awareness, so kind of looking at different types of data, app data, M2M data, IOT, mobile, all of this sort of coming together. And we are hearing constantly that the older infrastructure models are becoming obsolete. They're going by way of the dodo. And so really, people are looking at newer infrastructures. You're all here. Your colleagues are here to look at why and how we can implement the new next generation of infrastructure. And so we kind of started this research two years ago to really look at what is it that people are looking for. And so software-defined infrastructure, it doesn't really do justice to things like OpenStack or competing standards, but what it really is is a set of attributes that people are looking for. So from a platform perspective, certainly industry standards, elastic, open. From a value perspective, they want to be able to decouple the software from the hardware. And it's all about speed, efficiency, agility. Automation is certainly top of everyone's mind. They want to be able to go beyond just provisioning and configuration to an ongoing management paradigm that is totally different. And then finally, they want to be able to deliver the unified units of IT. We don't have to have silos of compute, network, storage. And really policy-driven, being able to have a unified service-level orchestration and service-level attributes that are tied to predefined policies and such. So key thing is open and elastic automation is the underlying principle for software-defined infrastructure. And a lot of those get carried over into when people are implementing OpenStack, these principles are alive and well. And so we talk to people who are in different stages of software-defined infrastructure. And clearly, test and dev and proof of concept is high on people's mind. But look at all the others. I mean, there's production less than six months, production six to 12 months, and production more than 12 months. And it jives with some of the survey results from the OpenStack Foundation that people are, in fact, moving SDI into layers of production. And they're cautiously moving there, but they're surely moving there. It is something that is eventually going to happen. Oops, not what this is. All right, so then how are people deploying software-defined infrastructure? So clearly heterogeneous applications is the most common use case. Private Cloud and infrastructure as a service is lower, but sort of really looking at applications. So from an application perspective, different types of projects make it to SDI a lot more quickly than a certain service as a service kind of a use case. And homogenous applications are a lot lower from that perspective. And so we talked with a whole lot of users on how and why and what of SDI. And these were the five key findings on software-defined infrastructure. The first one is it has to accelerate the time to market. The second is it has to deliver on the promise of automation. Third is cost reduction is certainly key. You have to be able to avoid vendor lock-in, so this is not something that could be a, so vendor lock-in. And then finally to increase agility in the environment. Sort of going back to the slide before, which is older infrastructure paradigms are becoming obsolete partly because they can't deliver on these five key things that we're finding out. So time to market. So we are constantly hearing from users who are basically saying, hey, my business wants me to respond more quickly. I don't have the luxury of implementing infrastructure in the old ways, which used to take days, weeks, months. And I want to be able to quickly deliver infrastructure units to the business so that they can develop applications or deploy applications. And the expectation here is that IT deliver their service at the same rate as public cloud. And I think the Boris from Mirantis summed it up very well where even his internal teams were willing to go to the public cloud if the internal IT folks were not able to deliver on their requests. So we have that option, many companies have that option and they're willing to use that option if their internal IT doesn't respond that well. So the second one is automation. So automation is really time to market, accelerating time to market, avoiding human errors and to abstract the complexity of the infrastructure. So it's really about taking units of infrastructure and delivering them in a way that is kind of automated. It's an acknowledgement that automation has to be end to end. It cannot just be in one place and be completely lacking in another. And it's really work in progress. So people will tell us that they are still not there yet when it comes to automation. They are still in the process of becoming fully automated. Now clearly the maturity level depends from company to company. Someone at a service provider level has probably got a whole lot more automation in their environment than someone who's just starting off on a journey for software defined infrastructure. Cost savings, again, it's about moving to an agile cost model where you don't have to provision storage or provision compute way ahead of time. And you wanna have both CAPEX savings as well as OPEC savings. And it's about savings from using standard hardware and then savings from using orchestration and automation so that you are using less resources in the process, both human resources as well as technical resources. The word commodity has become a bad word. A lot of times people prefer the term industry standard. Industry standard or commodity pick your term. The goal here or goal for folks is when you standardize your hardware irrespective of whether it is for compute storage or networking, you get those cost savings and those cost savings go across the board. They are horizontal in nature, not specific to any particular type of infrastructure. Vendor lock-in, so we heard loud and clear today that one of the people associate standard APIs with lack of vendor, I mean with vendor or no vendor lock-ins. The more standardization you have, the less the vendor lock-in. And once you standardize on hardware, even once you standardize on APIs, you're automatically minimizing vendor lock-in. Now, vendor lock-in and dependency is not so much a curse as much as it is really of sort of the opposite effect of choice, right? So you wanna be able to leverage, oops, you wanna be able to, sorry, I'm running on battery, that's probably why. So you wanna be able to leverage standard hardware and then use different software, purpose-built software that delivers different feature function. And what I mean by that, so for example, you might run Ceph on standard hardware, you might run Swift on standard hardware. You might also choose to run Scale.io or Nutanix or Nexenta or I'm just rambling out names here, but the point is they all run on standard hardware. So you can choose the stack that you want to run for delivering the right type of storage service as an example. Same thing with choice of hypervisor, same thing with networking. It's really about using as many industry standard components and then using purpose-built software with standard APIs so that it almost becomes a Lego brick approach. And then finally, agility, we want to be able to respond to business demands. We want to be able to automatically dial up or dial down based on resource requirements. So you might have on demand, if your business has a burst of demand from your business, then you want to be able to dial up your infrastructure. And then when you go back to the normal mode, you want to be able to dial down the infrastructure. So you want to have agility in your environment. So sort of summing it up, people are demanding, businesses are demanding a new infrastructure paradigm. You know, really it is about can my infrastructure scale on demand? Is it highly agile? Does it allow easy cost control? You know, it's sort of a shift from CapEx to OpEx. And is it service-based? Does it allow me to offer my infrastructure and hence higher levels as a service, as an internal service? And that is the promise. I don't think people are there yet. I don't think businesses are there yet. Clearly, it's sort of a meeting in the middle from second platform to third platform, which is really think of it as traditional IT and as Gardner puts it, bimodal IT. People are along the journey here and so is their infrastructure. Transformation is a process. And I think it's going to take a few months, if not years to really get there. And our research is showing that this is work in progress. We certainly are seeing a lot more traction, but we're not there yet. And this sort of sums it up in terms of how we see a software-defined infrastructure. And as the theme of this presentation says, this is really a proxy for OpenStack adoption. And OpenStack is one of the ways in which folks can implement a software-defined infrastructure as people call it software-defined data center. And therefore, morph themselves into being future-proof for applications of the future to deliver business in the future. And with that, I'm going to call my friend, colleague Larry. Okay, so now I get into the spotlight. So thanks for joining this session. We're going to talk about cloud and open source and what we have learned from some of our surveys. And just to give you the survey results, what we are seeing, and this is a bunch of questions we asked a large group last, about the December, January timeframe. And what we are seeing from overall is a high interest in open source in this slide. But what I want to now break this up because there are about 12 different questions that we asked them on different open source initiatives. If you look at what is the cloud management category, you might think that cloud stack is totally off, but people still are interested in that. That's one of the areas they care about. And in this slide, and I'll talk a little more later, that is we asked them a question of open stack with and without distribution, with support from somebody. So both of them are about the same, but obviously with support is a little more. So you need to look at that. Obviously there are different reasons that different customers look for support or not support, and depending on the type of the customer. I kind of categorize both of these as cloud platforms, both Docker and Cloud Foundry, also in a little different way, big interest in that from a group of customers. And then when you look at cloud orchestration, Chef, Puppet, Tosca, this slide, this looks like a little more when you look at an overall view, and I'm going to tie it with maturity of customers in a later slide. So this interest in continuous integration, continuous delivery is still quite high within customers. And then now when you go to data streaming, just we already understand the growth of data, the importance of real-time data, connected with automation, internet of things. This is not off the radar for most customers. And finally, when you look at integration, you talk about multi-cloud, hybrid cloud, that is another area of interest for customers. So I'm just going to go now, and when you look at it separately, you see there is a higher interest of cloud management, because we don't have some of the overlap between these areas. The cloud management, which is obviously open stack and cloud stack still dominates the interest from customers, especially on a private or on-premise environment, followed by cloud platforms and cloud orchestration, which is about the same. And then you've got data streaming and cloud integration. So these are again broken up by category. Now what we do within IDC when we do surveys is break up the customers and ask them, where are you on your cloud journey? Because comparing two customers, they're not all the same. So we look at, are they just starting in an exploratory environment or have they moved all the way? So we break it down into very clear ways, ask them detailed questions to see where they are in the cloud journey. Ad hoc, where they're just exploring or opportunistic, where they have now got into a collaborative and standardized environment, followed by now they have a repeatable strategy, they cover and meeting the needs of each group within the organization, or they're going all the way up to an industrial cloud, which could be what GE or Siemens and others are doing, building that, and they have actually made it into a practice within the organization. And otherwise going all the way into innovation and transformation of their fundamental strategy so that they can meet the disruptive challenges within the organization. So once we look at this, we took our responses and then mapped it up with them and here you'll see a very interesting difference. The ones who are more on the forward thinking or have already adopted cloud, managed and optimized, have a much higher interest in open source. So you can see that they now value the community improvement of a product rather than just buy it from one company but also shows that they have now collected the skills. They have understood what it is to use a cloud from one provider to now saying, I'm gonna go into an environment that is not going to be locked in, I'll be able to contribute my value to this and get value from a community rather than work with one company alone. This is something that you got to see from a maturity standpoint where our companies viewing this, they have now learned from their past experience and have made a decision to go with an open source strategy and collect from there. And now if you start breaking it up again, even by category, so on the left hand side you'll see just the ad hoc and opportunistic and on the right hand you'll see repeatable, managed and optimized. Here you'll see the cloud management much, much higher at 75% total from repeatable from what they are interested in open source. So whether they're getting it from a public or in a hybrid environment, these fall in different categories and all of these are of interest but you got to look at integration also is not of the list. Everywhere from integration to data to orchestration and platforms are all on the interest list of customers. Although we are here at this conference primarily talking cloud management, there are many aspects of cloud that you need to pay attention to. So I want to talk a little bit on the key takeaways and then leave some time for Q and A for us. So on the vendors, so I'm breaking this up into vendors, customers and standards organization. First of all, open source, open standards are playing a very important role. That's the overall view. But for vendors, I see that there's only one company really Red Hat who has really learned how to monetize the complete cloud open source and make money out of it. Other than that, a lot of customers are still in a combination of open source and proprietary software. So they need to look at what's going to be the impact on their monetization, on their business model and how are they going to make money off of it. And then if you're all on open source, what is your unique differentiator? And I see a lot of customers thinking, yes, we got portability, we are not locked in, but how do I really differentiate myself? Why am I different? And a lot of them are talking about the foundational aspects, their infrastructure or their network or there is some unique aspect, security and what do I do different? What do I bring to the table that others who are also using the same open source are not doing? So you're going to see that as an important area that customers are going to look at. As a result, I think vendors have to very clearly define why IBM versus HP who are both doing open stack versus somebody else. On the customer side, I think you'll heard on the previous presentations also talking about skills. And what I think from customers is they all have a lack of skills, not just on use of cloud technology, but the culture that has to be changed to try a cloud strategy and move in a cloud first direction. I live in Cincinnati and I talk to a lot of large companies over there and one of the things they tell me is their biggest difference between Silicon Valley born in the cloud companies and them is they don't try enough. And they have never had that as thinking that I can take a concept and try it out in two weeks or four weeks or six weeks and make a change. They typically are thinking of six months, eight months before they even find out whether they can do it or not. That's a cultural difference. And I think that's tied with what skills you bring and how do you go adopting cloud. The next one for customers is set a clear policy. What are you gonna do with open standards? What are you allowed to do? Are you going to use one with support or without support? Do you have the skills and the high-end skills that you need for adopting a non-supported cloud version? And that is a big difference and you got to make it very clear right from the get go. You may have seen a lot of strategy that GE has put out and they know exactly how they're going to use it. Not only that, but also when you adopt a cloud technology they have automated ways of seeing whether you adhere to the policy or not. Otherwise it'll take you forever before you start something up. And then going to the standards organizations like the foundation we are here with or even the cloud foundry or others. Do they have awareness of these standards of these capabilities that they're now bringing to the table? The biggest issue for customers is lock-in because they have been locked in before everywhere from databases to operating systems to hardware. And now they need to say that this understand that this is an approach they can take and prevent lock-in and that awareness has to be built within customers. And finally I see a lot of the open source foundations not involving customers enough. There is always vendors who are typically getting recognized for how much code they contributed but you got to also recognize the customers as to what they're doing even if they're providing you feedback and they're giving you code, contributing to the open source initiative, can you recognize them when they provide that? So I don't know how much time we have but I think we have a mic up there and if there is any questions, we'll be happy to take them or our contact information is up here. You can send us an email or contact us via Twitter or anything like that. Any questions? There are two mics up here I think. If not, I don't know what time we have. We have some time to give you back. Fantastic, or we've put you to sleep. But thank you for hanging in there. Thank you. Thanks.