 All right, how are you guys doing? All right, I heard a lot of these rooms are packed. So if you've got a seat by you, raise your hand. Maybe you could fit one more person by you. I think there's one right there. That's about it. All right, so my name's James Lobaki. I'm the product owner for Red Hat Cloud Infrastructure at Red Hat. This is actually my first presentation using Prezi. There's a big screen in a small room. So hopefully, if you have motion sickness problems, you might want to close your eyes during the transition. So what I want to do is take you guys through just a quick overview of what Red Hat's been doing in the cloud infrastructure space, then get into kind of a technical overview of the pieces. So I'm just curious, how many of you are like system administrators, operator-focused? Developers, partners. OK, cool. All right, so just to dive in real quick, I'd like to start out just by saying that Red Hat, these are just contributions. You could click the Stackletics link there if you want. I'll post these slides afterwards. We contribute a lot. We are committed to the OpenStack community. These are our contribution numbers here, just to give you an idea. And it's not only about contributing to OpenStack, but it's about contributing to every project. So I like to preface every presentation with this, that we are 100% upstream-first policy, right? So what's our vision? So at Red Hat, kind of our vision and our goal is to provide workload portability from an infrastructure standpoint across the four footprints that you usually see, physical, virtual, private, and public cloud. So we do this with both Red Hat Enterprise Linux running across there as a stable infrastructure. Private Cloud, obviously, provided by Red Hat Enterprise Linux OpenStack platform. But then we also do this by providing consistent storage footprint with Red Hat Storage, so whether that's Ceph or Gluster. And we also provide kind of seamless management with cloud forms as part of that. And then, of course, with JBossMiddle, where we're giving you application portability. So that's our main goal, kind of, as far as where we're going from an open hybrid cloud standpoint. You might have heard this morning, or if you saw the press release. So what we announced this morning was something called Red Hat Cloud Suite for Applications. Basically what this is, is this is an offering that we have an early access program now available. So if you want to run from virtual machines to cloud-enabled virtual machines on infrastructure as a service based on OpenStack, all the way up to containers running in a platform as a service with OpenShift Enterprise, along with hybrid management we're now providing us, just to give you an idea. So everything from fully open source from bare metal to apps. And what we're going to focus on today, so if you look at it really, what we built is kind of on top of a huge certified hardware ecosystem that we've inherited from building everything on top of Red Hat Enterprise Linux. We've built kind of the virtualization and infrastructure as a service. So in the virtualization space we leverage Red Hat Enterprise virtualization, which is, I'll get into the upstream communities that we work in here. And on infrastructure as a service, obviously we leverage OpenStack. But we also include hybrid management. And so what Red Hat Cloud Suite for Applications does is it builds on what Red Hat Cloud Infrastructure is an offering, and it adds platform as a service to it. So really what we're starting to do is build not only kind of a mode one, if you're familiar with kind of the Gardner terminology, mode one of traditional applications running as virtual machines, but we're also allowing people to run cloud-enabled virtual machines all the way up to container-based workloads in a platform as a service. And of course we provide kind of the middleware and mobility services on top of there through the middleware, the Jboss portfolio. So we do all this in communities. So we're here at OpenStack Summit, so everybody's interested in the OpenStack community. And you guys are all familiar with RDO. Who doesn't know what RDO is? Anybody? OK, it's Red Hat's OpenStack distribution. But there's a number of other projects that we bring together in our cloud infrastructure ranging from Forman, which becomes Red Hat Satellite, to Manage IQ, which becomes Red Hat Cloud Forms, and then a number of other projects like Kubernetes that we're leveraging inside of OpenShift, Project Tomic and Overt, which become Relatomic Coast and Rev. So everything is kind of this community enterprise model. What I'm going to dive into is really kind of this technical architecture. Don't worry, I'll zoom in on it. So basically the building blocks, again, of Red Hat Cloud infrastructure are really designed to allow you to run both mode one applications, so state full scale-up workloads, alongside stateless scale-out workloads with OpenStack. So the building blocks are basically Rev for Red Hat Enterprise Virtualization, our OpenStack platform, and then we leverage CloudForms for hybrid cloud management and Satellite for configuration and life cycle management. So I'll dive into these a little bit more. So on the OpenStack side, we build out Red Hat Enterprise Linux OpenStack platform on a huge hardware certification program, lots of partners. This is really built for your scale-out workloads. I'm not going to spend a lot of time on this because you guys are all here. You know what OpenStack is, presumably. But I think the key is that we also provide, for state scale-up workloads, we're providing Red Hat Enterprise Virtualization. And so this is for workloads that aren't going to fit onto your OpenStack environment. You can leverage Red Hat Enterprise Virtualization to provide kind of that stateful infrastructure. And then from a management standpoint, to provide kind of seamless management between these two platforms, and a number of other ones, which I'll get into, is Red Hat CloudForms. So CloudForms is based on an acquisition we did in 2012 of a company called ManageIQ. We've since open sourced all that code. So it's about $100 million acquisition we did in 2012. We've open sourced all that code in ManageIQ.org. What this is, it's a Ruby on Rails-based application that you can deploy as a virtual appliance, either into OpenStack, into Rev, into vSphere, so on and so forth, to basically start to begin providing a cloud management platform for that infrastructure. So you often hear with OpenStack, people kind of expect it to be an on-premise Amazon-like infrastructure. But the reality is that OpenStack provides an infrastructure platform. It's not providing necessarily the management platform. Things like charge back, full charge back, full show back, passing utilization collection, all those sorts of things. So it's built on a modern framework. It does basically passing utilization event collection, state relationship config, across a number of infrastructure platforms. And then it basically provides those up into a virtual machine database. And then from there you have a management interface and a northbound web services API that you can use. So it's leveraging Ruby here. So if you guys are interested, learning more about ManageIQ and I'll get through some of the use cases that we deliver through here. There's actually a ManageIQ day here that we have. So you can go there. There's a couple of labs. You can actually sit down, hands on the keyboard. These are running up in something called Revello, which runs on top of Amazon web services. So we have a full OpenStack environment with cloud forms that you can leverage for basically a lab and walk through all these different use cases of tenant workflows and service catalogs and all those pieces. All right. So on the satellite. So the satellite really is based on the Forman project. And really what it provides is consistent content and configuration management. So what we leverage satellite for in the context of our cloud infrastructure is mirroring content down and then providing a lifecycle around it. So whether it's dev, QE, prod, making sure that those are there. And the configuration management is all based on puppet. So actually you can leverage a puppet master inside of satellite and basically provide from dev to QE to production the ability to life cycle this, all your content. And that works across a number of areas. So dive into it a little bit more here. If you look at cloud forms, it provides not only consistent operations management on Reve and OpenStack, but it actually works across Hyper-V, vSphere, and AWS. So the key here is that you can drop this in. It doesn't need a green field environment. You don't need to deploy new infrastructure for this. You can literally drop the supply into your existing VMware environment, discover it, begin to discover the workloads there. And then you can optimize your VMware environment at the same time you start to plan out your OpenStack deployments. We have a number of customers that are using cloud forms actually to analyze their VMware environment and determine what workloads are best to move on to OpenStack. And so we have customers who are actually doing this migration today. All right. By the way, we can keep this interactive. If you have any questions, feel free to ask. There's a microphone in the middle, so there's a price to pay if you want to ask. You have to get up and make your way through everybody. All right, so on to satellite. So one of the things about satellite that is really crucial for this is basically providing that consistent content management. So whether it's syncing your configs from Git or pulling down certified content from our content delivery network or the ones you set up, it provides consistent life cycle management for not only the hosts, but also the guests. So whether it's your control nodes on OpenStack or it's your compute nodes or it's the instances running on there, satellite can provide consistency on how you mirror that content down, get it out there, and deploy it. And it can also update cloud forms as well. So the cloud forms, all this gets pulled together. So you don't need to have multiple tools, one for your hosts, one for your tenants, one for your management system. You can kind of unify that life cycle management between everything. And so what it also allows you to do is basically the combination of cloud forms and satellite with OpenStack or Rev or any of the other providers that Cloud Form supports can start to provide this kind of what we call standard operating environment and kind of consistent workflow. So it's great when you get OpenStack up and running, but then the question is how do you actually automate your tenant workloads on top of there? So just providing a QCOW image with cloud in it might be good enough for some workloads, but there's a lot of users that want to actually build out very complex applications on top of there. And so what cloud forms can do is it can actually help people basically with the automation engine can actually reach down into whether it's a control node on OpenStack or the manager on Rev or a V center on VMware or AWS. It can build up that entire instance using the native OpenStack APIs. Again, we're upstream only. This is nothing special we're doing here. And once it does that, it can actually do all the workflow and approval. So once your instance is built, then the instance can actually reach back out over to satellite. And that's kind of that step two there, where that satellite can handle all the provisioning. So whether it's what network cards you want on there, what IP addresses, what users, what RBAC control, role-based access controls, everything built onto that system the way you want it with controlled content. And then it could be handed back over to the automation engine. And this way your users are basically able to start to build out way more complex applications on top of OpenStack, Rev or the like. And just from CloudForms 3.2, which is based on, again, based on the managed IQ community, is actually going to support heat templates as well. So you can actually start to leverage all your heat templates inside of OpenStack for deployment there. So that's kind of the workflow automation here. Let me just also say that right now, the way up until version 3.2, which is the release is kind of imminent, CloudForms looked at OpenStack the same way it looked at AWS. So you'd come in as a tenant. All you would see is your tenant workloads. But we've done a lot of work in the Kilo release. Actually, Hugh, right there, so Hugh's team did a lot of work at Red Hat to basically instrument a lot of the projects inside of the Kilo release, Ironic, Tuscar and the like, to basically expose more information about the actual OpenStack infrastructure. So one of the big requests we got from enterprise customers was that they wanted to be able to see into that infrastructure layer of OpenStack and not just view it as a tenant, so that they could actually start to do capacity utilization planning, control policies, those sorts of things. And so in the Kilo release of OpenStack, we've instrumented all that into the kind of control plane, right, of the overcloud. And so CloudForms can actually add an infrastructure provider now and look at OpenStack as an infrastructure provider. So we're able to start pulling out a lot more metrics around CPU memory utilization, all those sorts of things. I think you're going to see, over the coming weeks, you're going to see some pretty cool demonstrations we do where we actually hook CloudForms, actually manage IQ, because we'll probably use the upstream code to demonstrate this. Where manage IQ is actually looking at our RDO, our OpenStack distribution, and being able to automatically scale this based on capacity utilization data. So pretty cool stuff. Oh, is he? OK, who's that, John? Yeah, so we've got a whole bunch of sessions that I'll plug at the end, and that's probably going to be demonstrated here. All right, so just from a conceptual standpoint, because I know everybody here, we all love OpenStack, right? I love OpenStack, but the reality is that there's probably some, strategically, there's a reason people want to separate their cloud management platform from their infrastructure platform. And part of it is that you don't want to push too much business logic into your infrastructure platform. So by leveraging a cloud management platform, what it allows you to do is you can gain a service catalog that works across providers. And things like governance quotas, approvals, and placement logic can be separated from the actual infrastructure platform. And then you can leverage whatever infrastructure platform you want, whether it's a heat template, a VMware template, Rev, Pixie, whatever you want from your infrastructure provisioning, you can leverage that. And then on the other end of that, you can also decouple yourself from the infrastructure by leveraging systems management. So in this case with Forman, you could leverage things like content views and host groups, so you can build automation that works across all your infrastructure platform, and it's not just tied to a single silo of OpenStack. All right. So three things I want to touch on. So between Rev and OpenStack, we've been doing a lot of work in bringing together networking, identity management, and storage. So you can have a common fabric between both of these. So this is pretty interesting work because what it allows you to do is simplify your architecture so that you can be leveraging neutron across mode one and mode two, as well as identity management. So on the identity management side, I told you I'd give you guys motion sickness. All right. So on the identity management side, what we've been doing is leveraging Red Hat's identity management, which is based on something called IPA. And so what IPA allows you to do is actually allows you to do cross domain trusts, even with Active Directory. And so you can start to build a unified single sign-on between all these. So for example, with OpenStack, we've done a lot of work so that Kerberos and SSO works seamlessly. So when users log into OpenStack, they can also log into Satellite, CloudForms, and Rev with that same identity. And it flows all across. And again, this is kind of simplifying the architecture, making it more usable so you don't have to log in multiple times, have multiple users, and multiple places. On the neutron side, this is actually new in Rev 3.5. What we've actually done is there's something called the Rev Manager in Red Hat Enterprise Virtualization. You can think of it like vCenter or Microsoft System Center, Virtual Machine Manager. And it can actually look and discover neutron networks in your OpenStack environment. So if I have neutron networks in OpenStack, my OpenStack distribution, I can actually discover them with Rev Manager, and I can make them available to my virtual machines running on Rev. So now I can start to build a common tenant network between mode one and mode two. This is kind of simplifying all those pieces between here. So I don't have to learn one networking technology and how to provision a network for my virtualization and one for OpenStack. And then finally, on the storage side, we support glance between the two. So basically, you could dump all your images and templates in one spot. And then you could leverage these images and templates across both Rev and OpenStack. So we're kind of combining these two areas. In the 3.6 release of Rev, we're also adding Cinder support for this as well in Rev. So basically, you'll be able to use Cinder in Rev to back all your stateful virtual machines as well. All right, no questions. I should have brought some hats. Hats always get questions. All right, so just to give you guys an overview of where these kind of capabilities lie, a lot of people say they want to deliver AWS in their data center. The reality is that between OpenStack and Rev, what you're getting is mostly solving the problem of compute, persistent storage, object storage. Don't get me wrong, these are important problems. The same problems Rev solves as well just in a different paradigm. But then when you move into areas of analytics, monitoring, IT service catalogs, charge back, those pieces are really solved by cloud forms and the cloud infrastructure play. And then satellites adding onto that kind of tenant provisioning, configuration management, software distribution as well. All right, so two things I want to kind of, one other thing I want to kind of impart on you guys is this idea of how to deploy. So we kind of view, at the infrastructure layer, you have two options, right? You can kind of nest your virtual infrastructure, your existing virtual infrastructure underneath OpenStack, right? And you probably see people talking about this even here. But from our standpoint, we feel that on the left-hand side, you kind of lose the ability to do centralized operations management when you do that. And second, you kind of pin yourself to a scale up architecture underneath OpenStack, which is kind of fundamentally flawed, right? OpenStack is meant to scale out. And so the way we actually work on our deployments is we leverage cloud forms as kind of a broker between vSphere, OpenStack, Rev, Hyper-V, AWS so that users can get a consistent experience with that. But really, we leverage KVM underneath OpenStack as opposed to trying to kind of shoehorn a vertically scaling architecture underneath a scale out architecture. All right. So a couple of options you have here. So you can actually coexist your VM work footprint with either Red Hat Enterprise Virtualization or OpenStack. I'm going to actually skip this one and just go right to OpenStack. So basically, we've had a number of people that are actually interested in migrating from VM or vSphere over to OpenStack. And so we are doing this today. So there's kind of two ways of doing this. The first is leveraging something like cloud forms and actually doing a vert v2v copy. So actually, you see number one there. Basically, you're taking the virtual machine. You're using the vert v2v tooling that exists in Red Hat Enterprise Linux. You are converting that virtual machine over. And that works. It's a binary operation. It's very expensive. But we have users who are doing this today at scale. The second option that we see probably more people doing is that they are using the cloud management platform and then brokering new workloads onto OpenStack and some workloads onto vSphere based on tagging, capacity utilization, what the application is. So this is probably a far more common request that we have. And so that's really where cloud forms shine. So it basically builds workflow and decision making into the cloud management platform. And then based on the person's user ID or tagging or however you want to build your taxonomy inside there, it can decide where it's going to place a workload. So it might be that a developer logs in and you determine that developer can go onto OpenStack. And then when a production person logs in to deploy an application, that's going to get dropped onto OpenStack. So we see a lot more interest in that kind of approach. All right. So just to give you an idea of some of the common management integration. So we integrate in two different places. One is on the management side. And here we integrate with almost every popular configuration management system. No CF engine on there. Sorry. But CMDBs, catalogs, workloads, infrastructure. So whether it's Puppet, Chef, Ansible, we have numerous customers doing all of these kinds of integrations today. Or whether it's integrating with your IP address management systems like Infoblox. And then satellite actually integrates with Git and Jenkins so you could do continuous delivery of your kind of configuration data as well. And then on the infrastructure side, this is just a sampling of them. This is probably one of the biggest, I think, one of the largest IHV independent hardware vendor certification programs in the industry through Red Hat. And we extend this across both Red Hat Enterprise virtualization and our OpenStack platform. And the reason we could do this is because it's all based on Red Hat Enterprise Linux. And so we deliver this huge ecosystem around there. So whether it's storage, network, or compute, you have kind of a huge choice there. So any questions? You don't have to use the microphone. Sorry. All right. So what I'll do is I'll let you know about a couple of sessions here. We have a bunch of sessions going on. I will definitely tell you the managed IQ sessions going to be great on Wednesday. Definitely attend that. There's a whole bunch of live labs there. We have the other one I was going to mention is Mark McLaughlin's keynote tomorrow morning. But with that, I'll take any questions. Hi, James. Thank you. Last year, we looked at managed IQ, and we found basically two areas where we found it wanting. One was single tenant. It was very much single tenant oriented. So I think we're going to have a look at 3.2 now. The other one was the integration with the solometer, which we found extremely expensive. Solometer itself was not without its problems. Can you say something about solometer and managed IQ integration today? Yes, I absolutely can. But even better would be Zav. And you can talk more about it. All right. So there's a lot of multi-tenancy work happening right now upstream in managed IQ. The target, so right now where we're at, is CloudForms 3.2 is going to be the next release. Beyond that, the six-month release cadence we're on, the plan is to get true multi-tenancy done in that six-month window. Around solometer, there's been a number of improvements around solometer data gathering in the 3.2 release as well, not only on the managed IQ side, but on solometer itself. And so those I think you'll see in the 3.2 release. OK, thank you. One last question. CloudForms also talks to Microsoft VMM, doesn't it? Yeah, so in the 3.2 time frame, that is the goal. So that release is kind of imminent. So that could be another way to get out of Microsoft VMM, actually, the same way you can get out of VMware. Thank you for saying that. I didn't say it. You're welcome. Can I get a hat now? Yes, I don't have any hats. But if you stop by the booth later, no. All right. There was not a plant, by the way. I promise. Anybody else? All right. Well, thanks, guys. I'll give you back some time, I guess. All right.