 All right, thanks for coming everybody. So I'm gonna go ahead and get started. I think it's 1115 My name is James Lobaki. I'm a product manager at red hat working on something called red hat cloud infrastructure Which is a combination of technologies is open stack along with red hat enterprise virtualization There's our data center virtualization software and cloud management platform called cloud forms And recently we just announced this week the new version of our HCI our infrastructure offering Which now includes red hat satellite our systems management portfolio as well, but what I want to do today is talk about kind of this the significance of open clouds, so There's really four Main shifts that we see taking place In IT architectures. There's a shift taking place in the infrastructure a shift taking place in the platform a Shift taking place in business and middle-war services and then finally an IT operations management And these are four areas that red hat really is focused on As you know that we're focused on when these shifts are taking place So real quick from an infrastructure standpoint the shift that's taking place you guys are probably all aware of if you're here, right? This is this is open stacks mission. It's basically transforming infrastructure, which is vertically scaling Using specialized storage and networking gear and maybe virtualization To a horizontally scaling infrastructure as a service platform like open stack So this is whether it's operating system with compute software defined networking and software defined storage everything's defined Defined in storage and rather than scaling up with expensive proprietary Solutions with infrastructure you're scaling out on commodity hardware So the benefit is is pretty obvious with this right while you're moving to open stack The reason is is you're moving away from specialized hardware to commodity But then even one of the bigger benefits is that everything being software defined it allows you to automate a lot more So instead of having to you know Work with someone or have some sort of process in place that's declarative you could start to build out Kind of imperative structures on the right-hand side With infrastructure service, that's the first shift taking place At the same time Beyond just the world of open stack. There's a shift taking place in the platform space And so this is happening where people are moving from platforms, which you see on the left-hand side To really platform as a service So in that in the traditional way of doing things a platform was essentially You know a collection of operating systems with application servers on top of them and then application code on them So if I wanted to deploy a platform I would call up my admin and then he would the sys admin hand off to the application admin He would install the application server. He would configure it maybe for clustering all that would happen It was very bespoke. It was customized. It can only be done, you know by a certain individual and There were ways to automate that With configuration management languages and things like that, but they were Expensive to maintain that that configuration management and that life cycle management tooling on the left on the right What's happening with application platform as a service is two things First we are moving from a kind of imperative way of describing how you do things on the left-hand side To a declarative way on the right. So we have a pattern based way that you could do things Describe your application platform. So how for example, I might declare how I how my middleware and my database server relate And when I do on deployment time those two things can actually be merged together Without an administrator being involved or without some configuration management language that's or workflow process taking place And the second thing that's really leading the way here is containerization So containers are being used in in place of virtual machines. So that's driving up density So what this is also doing is It's also allowing you to just like infrastructure as a service. It's allowing since everything is defined in software It's it's allowing you to have programmatic access to this So no longer do I have to call one API in my virtualization layer and then call another API for my config management layer And then call another you know API for this I'm calling the application platform as a service and asking it for an application server. It's giving it to me Right. I get my application server. I get my source control repository and as a developer I don't have to worry about anything else. I pushed my source control repository and my applications are updated So that's that's a big advantage The the third shift is really in the business and middleware services space And so this kind of builds upon the application platform as a service So where the application platform as a service is a developer asking for an application server He might say I want to build a Java app and we give him, you know a Tomcat application server Customize the way he needs it and it connects to his source control repository for automatic, you know updates with Xpaz X platform as a service what we're really doing is we're taking Higher-level services like business process management, you know as an example or data virtualization And we're moving those into the paths. So now instead of simply asking for Ruby or Python and getting those pieces I'm asking for business process management, and I'm getting that so I no longer have to just get an application server I can actually get all this business value right out of the gate So even more automation That's the third third area Finally the fourth area is around the transition from IT operations management tools to a cloud management platform So you're probably all familiar with kind of the IT operations management tool set of old old CMDBs and service management and charge back and service desk capacity management Those were typically all from other vendors, right? So separate vendors for each one of those That reached out into your infrastructure to perform these tasks And what's happening is people are moving to a cloud management platform and the reason they're doing that is because a cloud management platform is able to iterate faster and Plugs into your different infrastructure service platform as a service and software as a service models And can keep up with their rate of change. So for example when open stack iterates today You know, how many of you have your your plug-ins from from your, you know old Change management systems that automatically plug into open stack very easily You don't right? It's very difficult to get that in place Well with a cloud management platform, there's not as much cruft around there and it's not as much legacy technology So those cloud management platforms are iterating faster and they're able to provide features into those platforms much faster So when you combine those for Architectures those shifts in architecture the shift in infrastructure to infrastructure as a service the shift in platform to platform as a service and the shift in business and prop business Business services to X paths and finally with a cloud management platform You really get the architecture for an open cloud and so it starts with Infrastructure as a service at the bottom software defined horizontally scaling It moves up into the application platform space where you can containerize your applications and also have predefined Tools for your developers to use and then it moves all the way up into X paths where you can get entire sets of services there And it's all pulled together with a cloud management platform So this is what we see taking place today people moving to this new architecture But it's it's not just about doing it in your own data center And that's something that's why an open cloud is so significant because if you're not building this with all open source technologies With a broad community Then what you're gonna end up with is you know another silo in your organization You're not solving the fundamental problem of how do you also move this out to be a hybrid cloud? So customers not only want this new architecture with all its efficiencies, but they want Two things they want to be able to use public and third-party providers So I'm gonna be able to move out to broker their workloads on to whether it's a public infrastructure as a service cloud like Amazon EC2 Or whether it's you know moving out to a public pass They want to be able to hosted paths such as Fred hats OpenShift hosted hosted platform. They want to have choice there They also want to be able to integrate into their existing systems So they have these other IT operations management tool sets that they don't want to get rid of They want to actually plug their cloud management platform into there and let their let their business take place So what about the workloads? So the Ark, you know, we probably hear all week like all the sessions here at OpenStack Summit And when I look at the majority of them It's all about how do you build your OpenStack cloud, right? And then you get it running nobody ever talks about the workloads, right? Like isn't that kind of the whole point like I think sometimes we we sit there We go wow, we got it installed and then we're like Okay, what does it actually do right? So if we don't actually focus on the workloads. It's it's not good So there's there's all these benefits that this new architecture whether it's infrastructure as a service or pass or XPAS bring and Here's really four of them one is you get much faster instantiation. So, you know, you can get it on demand whether it's a The ability to just access it through a self-service portal faster or whether it's spinning it up faster as a VM instance Or even as a container with which launches in you know a second or sub second launch time Or whether it's another benefit is autoscaling So you can actually autoscale these resources with you know perceived infinite elasticity The third is application recovery So since these services whether it's the infrastructure as a service platform or the application platform as a service Provide API's and kind of health monitoring whether it's heat in OpenStack or it's you know Some of the tooling in OpenShift or it's Kubernetes From a from a PAS layer. There's tooling in there that allows you to begin to autoscale your application And also auto recover it and the fourth is greater portability You can you can more easily move your workloads from on-premise systems to off-premise systems So those are really the four areas that these new methods Provide you so we're seeing when we see customers adopt This is a common theme that we see happening when when they're looking at their workloads And how do they move their workloads to this new model? This is kind of the pattern we see taking place So the first is really they go and they discover their patterns in there in their existing environment So what are my customers deploying right so they go out and they say what's in my data center today? And what are people requesting for me today? So they'll go and they'll look at that and this typically falls in four layers as you see So the the first step is really they look and they say okay, you know Maybe I'm maybe I'm automating the installation of my operating system, right? And really this takes a from an operating system view up the things around that the the actual machine or instance Those are you know kind of assumed here, but those are also important The second is they look at how they configure that so once I have an actual operating system running there How do I configure it with packages files and services? What do I have to do to that system? The third is how are they orchestrating this so are they? Just phoning up a system administrator and asking them to make changes Are they using a configuration management language like chef or pop it or ansible and then the fourth is how do they define their application layout? So if I'm launching an application server, how do I relate that application server to its database if it's an n-tier application? So on and so forth So that's the first thing they discover all the patterns in their data center And then what they do is they realize they start to realize Which ones are being requested most so they might they might figure out that you know? 80% of their developers are really just asking for a single you know single platform right maybe it's Tomcat or Ruby or Python that they're asking for and so they'll what what'll happen is they'll build a blueprint using workflow so Recognizing that they're an existing investment You know I let me ask is how many of you guys are running VMware inside your organization today? It's curious Okay, so a lot of you right so it's not expected that you would go and toss out all your VMware Infrastructure today right or all your physical infrastructure today or you know for that matter maybe even your mainframe right so It's not it's not expected that you would throw that out so the first step is that they automate with workflow and so what this means is they're using a Like a state machine based automation engine that goes out and does automation so maybe this is my user comes in to the self-service catalog and requests their their development stack and My automation engine basically goes out and it will it will touch the f5 load balancer and create a you know Create a new IP address for it add it to the pool It'll go out to VMware. It'll call down to vCenter. It'll create the VM right you see you probably all aware of this Right, so then then it'll go down to the net app. It'll it'll create a new volume on the net app for me It'll connect it to the VM. It'll do all those things right? So it's really the first step in data center automation. It reduces the manual processes, right? So and it also provides self-service to actually start to capture the users that are coming to the platform So now I actually know what my users are consuming and I'm doing this using my existing investments. I'm not throwing anything away But finally the problem with the workflow based method is that it's expensive when you update your your f5 load balancer Your workflow breaks right when you update vCenter your workflow right API changes, right? There's no compatibility between those necessarily and since the workflow has to touch all these different products from all these different vendors It's very difficult to align that workflow and keep that automated So what happens is they begin to add these cloud platform patterns to the blueprint So a great example of this is heat how many you guys are familiar with heat or our views Okay, so you know heats great So why would I go out and if I if I have heat available? Why would I go out and? provision my instance and then provision my Provision my instance for web tier and then provision my database instance and then tie them together manually when I could leverage heat And I know that from version to version heat is going to remain you know somewhat somewhat compatible, right? So in that way I get I get away from using workflows Which will break between versions and I move to heat which is a native capability of the platform, right? That's what I want to move to I want to move away from workflow and towards the pattern so This is really kind of the evolution from this data center automation to this cloud automation And it doesn't have to be a rip and replace or a new silo you can really build this into there So you get great of native native functionality So things like auto scaling and heat or application recovery that you've seen maybe from some demonstrations this week You can do that and it really provides greater resilience All right, so what does this look like so we talked about the traditional platforms before So if on the left-hand side, this was your traditional platform, you know, maybe this is your You know you got some some net app or emc storage, you know your storage frame here You've got VMs running on top of this specialized hardware with some operating systems on top and Database Python and web tier all tied together, right? And you've got that running when your developer comes in and he asks for that and he says I'm starting a new development a new development a new application Maybe you're literally like procuring all this hardware in Iraq and doing this That's like the the Stone Age way and and perhaps you know people are still doing that, right? But you want to move on to this new platform over here, so how do you do that? Well, the first thing is you go out and you discover that pattern, right? You go and you say how are they how are they doing this? What what network equipment are they using? What storage equipment are they using? How are they provisioning the operating system? How are they getting the application server on there? How are they doing all this once you figure that out? You basically create a blueprint in the cloud management platform. That's using workflow So you go out you say sorry about the the small type here We'll get the resolution higher next time So you got you create this blueprint that you then advertise out to your users You say if you want this stack this way on the right-hand side, maybe that's on VM Where maybe it's on red hat enterprise of virtualization. Maybe it's on hyper V If you want this stack where you can get your entire development environment come to our self-service catalog now select it and we'll automate it So when they do The requests start coming in and you can automatically provision that you're using workflow So the automation engine goes out to your f5 it lights it up goes out to the net app Does that lights up the VM talks to puppet or chef and makes all the magic happen, right? I picked that up from the keynote yesterday make the magic happen right if you guys were there all right So so then the user is already coming to your catalog So you've captured their usage and now what you can do is you can start to optimize so you can start to evolve to the cloud Platform so you could say okay now, you know, you probably want to notify your user You don't want to do it without them knowing right but you could you could then send and send an email Off them say hey mr. User next time you guys request your new development platform Your database and Python layer those are still gonna be running in VMs They're gonna be running in a virtual machine, but they're gonna be on open stack now, right? And so suddenly you're moving them over to the lower cost platform And maybe you're actually starting to define that with heat and Then finally you start to move them over each layer by layer So then you look at their application layer and you say well you're using a database in Python Well, it's very easy for me to say I can go into your platform as a service layer And I can create two Docker images for that and I can create a Kubernetes template for that I could begin to construct that Platform as a service description. So the benefit here is One I've moved them from expensive specialized hardware onto infrastructure as a service already and everything is also software defined So now instead of making 12 API calls from cloud from my cloud management platform Over to the f5 and the net app and vCenter and rev or hyper v and all this stuff and all those other pieces I'm making one call to open stack and it could define that you know, it's one API, right? With the platform as a service, I'm starting to get massive density So on this slide, we're showing you know a couple of containers running on here The reality is is that we've we've seen like you know a thousand applications running on a single operating system In our hosted platform as a service. So think massive density. So Reduce the number of operating systems I have running. Yes This is this is red hat telling you to reduce the number of operating systems you have running, right? So now you have this containerized application running here Very efficient and it's all software to find there as well and not only that but it's not a Customized pattern. I know it's a known quantity. I've given that to my developer. I didn't say Mr. Developer, please tell me how you want Tomcat configured or Ruby configured or Python configured I gave it to him and said this is the Python you're gonna be using here. It is right so another benefit and Then finally I can move actually there. They're you know, they're even higher level services into this X platform as a service, right? So I could begin saying not only are you gonna use Tomcat? But next time you want a mobile platform, I'm gonna give you the entire mobile platform, right? You're not gonna be using you're not just gonna be getting an application server and then deciding that so Basically, you're drawing a chalk. You're kind of snapping a chalk line on your infrastructure higher and higher To where more of it is standardized and less of it is controlled by the developer And the developers only doing what they care about all right So and by the way every time you're doing this you'll notice that this blueprint is basically updating So I'm basically replacing inside my workflow I'm taking huge sections of complex workflow out of my automation engine and I'm replacing them with those declarative languages in here So whether that's heat, you know instead of having you know Very complex workflow going out and touching all these pieces. I just replace that all with heat, right very simple All right So along with that it's important to provide lifecycle management so Part of this is being able to plug that into a development test and production framework so that you can actually Move when you're moving those applications They have access to the content they need and you can actually Make sure that you have the right the right content in the right place and The other benefit of this is as you move to those cloud-based patterns. They become more portable So whether it's being able to share it whether having your vendor provide these things such as you know heat templates or You know puppet manifests and those things or being able to author them yourself and have them in your own organizational repository And then being able to deploy them on your on your infrastructure So this is this is all not possible without having kind of a basis of an open cloud Right where everything in your entire portfolio is open source So this is where you're supposed to clap because I didn't mention any products until now and this is a vendor sponsored session so So this is this is essentially if you were to map red hats portfolio onto that open cloud This is how it maps out. So the basis of our infrastructure as a service is open stack It's red hat enterprise Linux open stack platform. So that provide and that's all built on red hat enterprise Linux horizontally scaling You know infrastructure Above that the application platform as a service that we were talking about is open shift So that's built on red hat enterprise Linux again and provides Containerization for your applications along with standardization and automation for the developer experience And then finally up at the top you have XPAS Which is our offering to basically take all of our business in middleware portfolio and put them into standardized services that you can consume And then on the right this all gets tied together through through cloud form So, you know that provides all the IT operations management functions for those different layers and Inside the blueprint That's basically what you're seeing here is a lot of this is provided by something called red hat satellite Which provides life cycle management for operating systems and configuration management But then you also have technologies like heat that are providing a topology description for your infrastructure As well as the open shift cartridge construct, which provides a topology description for your application layouts All right And even the best part right so the best part I would Urge anybody to throw this slide up there from any other organization in the world Completely open source right so the coolest thing about our portfolio is we're not the only ones building it right? We have open source communities for every one of our projects It's not that we say our infrastructure as a service is based on open stack and our management is proprietary or Vice versa We're not trying to say our management is open source and our infrastructure as a service is proprietary It's completely open right so every one of these is based on an open source project Whether it's satellite, which is based on the upstream of the foreman project Red Hat Enterprise Linux opens that platform obviously based on RDO our our Community distribution or open shift which has open shift origin community as well as cloud forms Which we actually announced the open sourcing of this year So we actually took you know hundred hundred million dollar acquisition that we made in 2012 and open source the entire code base and Our you know We just had the first design summit session on that and it's really kicking off with a bunch of partners and and users So completely open so you're not really getting innovation from one company So again, I think when you when you think about this and it certainly has the the the products and the solutions We have here, but the way I would I would suggest we look at things is where are you on kind of this the journey of these different architectures? Have you adopted cloud management? Are you at the infrastructure as a service layer? Are you at the platform as a service layer? Are you at XPAS and then if you are? Kind of what are you doing in your workload space? Have you done discovery against your workloads? Do you know what you have in-house? What's the most requested application that your dev teams use? What are they asking for? What are they developing on? What are they moving towards? Are you are have you implemented workflow? So have you automated your existing infrastructure if you haven't there's probably a huge you know return on on that alone And then third are you developing these patterns for the future where you're looking at? How do I develop a pattern for how to deploy all my virtual machines and instances? How do I develop a pattern for how many deploy containerized applications so on and so forth? So again Red Hat has a broad range of products and offerings for this But what I would urge you is think about start thinking this way like have you discovered have you built a workflow? Have you built patterns? All right So around community involvement, so I mentioned all of our all of our Red Hat's portfolio is open source I'm not going to get into this but too deep. Let me see Time check so I don't want to get in this too deep, but You know we have a long and you know history of open source contribution, right? We started contributing pretty quietly actually to open stack Mark McLaughlin who was on the keynote yesterday Was one of our first contributors, so he started with Nova and Oslo. I think in the in the in 2011 So I think that was not fulsome. It was Yeah, Essex right so he started in Essex in 2011 Again, we've been a you know number one contributor It's not it's not to say that being number one contributor is the most important thing the key takeaway is Being involved heavily in the in the open source community is very important Red Hat's known this from the Linux kernel days where you know if you need a kernel patch being able to you know Call the call the right people and talk to the right people and influence them is very important and We have kind of this this this approach to open stack where we're fully in committed to the open source community Right, but we also realize that it's kind of it's an unstable community distribution, right people use it for development Then we have RDO which is our you know bleeding edge. It's upstream It's but what we're doing is we're taking the source for packaging it as rpm so people could use it easily And of course we have enterprise Linux distros here, whether it's sent OS rel or fedora No support six month life cycle, so it's very fast moving and this is the same thing we did with with with Linux, right? We basically looked at Linux and we figured out Pretty early on that you know a lot of the investment banks and customers that were early adopters of Linux They loved it because of the innovation, but they hated it because of the maintenance And so you've been hearing a lot of debate about that In recent weeks about the the life cycle of open stack And so that's what red hat enterprise Linux open stack platform really addresses And that's the three-year life cycle with all the support and the ecosystem of certifications, so Basically all the certified ISVs and IHVs that can run on red hat enterprise Linux Open stack platform. All right, so that last bit here, so just to you know, this was shown to me this morning So the the if you took the actual if you go out to stack a little stack litics Which red hat doesn't control these statistics around stack litics If you look at the stats If you actually take the ice house release And you take the in advance and ink tank acquisitions that red hat did so Seth and the in advance folks We are actually the number one contributor again, not that it's Important about being number one, but it is important that we're heavily involved in investing So I'm not gonna jump into here. Let's skip over that, but I do want to tell you a couple things so Couple resources if you want to find out, you know, if you're ready for open stack or learn more about our hybrid cloud offerings There's the URLs. There's also a limited time offer. I was gonna say I was gonna do kind of an infomercial as a joke to it So there's a there's a if you buy your open stack certification exam you get the online learning course for free So definitely take advantage of this if you're interested in open stack training You can go down to our booth the first 500 to buy the exam get the related course for free for 90 days And you can guarantee your seat the first 10 to register with the promo code Prep for stack will be guaranteed a spot So hopefully by the time you guys are watching this on the video those have already been used. All right, and then Couple things on the additional sessions We've got a number of different sessions the authors of the open stack design guide is really cool So Vinny and Steve are putting that on they've basically written a design guide for open stack So if you want to talk to the guys who are you know knee deep and elbow deep and an open stack That's the way to do it The Neutron network node ha is definitely gonna be a great talk Nick Bar set from in advance Transforming to open stack really good talk Chris Wright on the opens he's a open daylight The lead of open daylight at red hat. He's giving a talk as well, and then mark is doing a talk on Oslo messaging I would say the one last piece too is the docker all the open stack services that Todd and I are doing or sorry Brent holding the nire doing tomorrow morning We're gonna give a demo of basically launching all the docker services All the open stack services using docker and kubernetes on something called red-hander prize Linux atomic In like two minutes. So basically from no open stack to open stack in two minutes. It's pretty cool So much more technical than this talk. So but if you're interested, please please stop by We have a big crew coming So let me with that. I'll take any any questions and thanks for Thanks for coming any questions No quiet crowd All right, thank you guys