 Good morning. Shall we get started? Let's go. My name is Amir Sharif. I lead the product or rather the partner integration team for Nuage Networks. Nuage Networks is an sdn company, and we work very closely with The open-stack community. Here, I want to talk about a couple of reference Architectures that we debuted yesterday with our partners Here on the show floor. But first, let me give you a little bit of overview about Our platform. So we have an sdn controller that's been derived from a Controller that's running on 40 of the top 40 service providers Around the world. It's the same controller that alcatel uses in their carrier Grade routers. And we have virtualized it, and we have a three-tier Hierarchy with policy, control, and basically the routers And switches inside the high provisors. These components have a lot of interfaces for bringing an Ecosystem of partners together, and I want to give you the Framework that we use for these reference architectures. So the extensibility framework starts with the cloud Consumption layer, or the set of apis that plug into what We call virtualized services directory to expose a set of Accessible apis for incorporation into cloud management Systems like OpenStack. We also have just announced a virtualized services Assurance platform, which is a way to correlate the Overlay network, or the virtualized network with the Underlay, which is the physical network. As it happens, things move around in the virtualized World, and if you have a problem running around in the VM somewhere in your data center, you want to figure Out where it is on the physical layer that the problem Happens, and you want to correlate those two events. Third, we have a set of virtual routing and switching apis, And this is where we incorporate with the physical layer on The network, so partners like orista, with whom we're Working very closely, are among the people that use These, consumes these apis and helps us expand the Ecosystem. And finally, we have an extensibility and scalability Framework that we use for service chaining, and this is where The value app comes in into the sdn layers. So the partners that we're working with, such as f5 for Acd, or paulowalta networks, they use extensibility and Security framework for rolling out the services. So this is all great and good and great. You might want to say, you know, are these test tube Experiments, or are they real? So let me talk briefly about what we have actually done To date. We started shipping via virtual services platform in 2013. In fact, the gentleman sitting in the audience was the Responsible product manager that shipped the platform. It's currently operational at very large data centers, and China telecom, it's scheduled to scale to 12 data centers With, let's just say, tens of thousands of vms, or tens Of thousands of hosts. So it's operational at a very large scale, and it's Installed in north america, europe, and asia. Basically, because we've been operational, we have been asked By customers to incorporate with other partners. So the customer has a particular requirement for a Distributed firewall, or particular requirement for an Existing firewall, and because of the flexibility of The platform, we've integrated with all of them. So in that sense, we've learned a few things. And we're basically encapsulating what we've learned into These reference architectures that i'm going to talk about Today. But before doing that, let me talk about the partners that We have at the various layers. As you see, we have had to work with a number of Partners, and it's been quite a good learning Experience, so with red hat, hb, canonical, Merantis, and so forth. So if we have a rich ecosystem that provides a rich set of Platforms. So let's talk about the reference architectures Themselves, and forgive me, i'm going pretty fast because There's only 20 minutes on the stage. But if you look at open stack, and the framework that it Has, neutron offers you core capabilities, and allows You to extend those capabilities, and this is what We're taking advantage of. So first, at the open stack layer for the first reference Architecture, we're using red hat open stack platform six, Or osv six. So with that, we then inject rsdn layer into the Extensions of neutron, and we have partnered with f5 for Load balancer as a service, and powel to networks for Firewall as a service, and finally with info blocks for Cp, dns, and itam. All that is running as an integrated service, and you can View this as a complete networking solution based on some of The leading companies in the networking space to offer very Good services for the data center, whether they're inside The enterprise or inside the service provider. We've also worked with canonical. That's a second reference architecture that we have, And we're using canonicals open stack distribution, and Again, we're injecting nuage networks as extension for Networking, and working with avi networks for a distributed load Balancer as a service. So avi is a startup in sunday bale, california, that has The ability to distribute the load balancer on a per Hypervisor basis, and it's a very clever solution. For firewall as a service, and the extensions like dhcp and Dns as a combination of nuage networks and native open stack Capability. So that's a second reference architecture that has come about That we're rolling around, and again, these are based on Existing operational models that we have rolled out. So a word about the physical layout of what we've done. So first, we have two physical servers, one for the Services directory where the policy layer runs and one for The controller. Then we extended the setup with four other physical servers, And the red hat osv6 jr release is running on those four. So you have the open stack controller, and you have what we Call the open stack vsrg, or open stack compute that's Running vsrg on there. Vsrg is our virtual routing and switching gateway which Provides the interface between the underlay network and the Overlay network. So it's what puts the vxlan tag on there to virtualize the Network and strips it off as you need to talk with physical Assets. Then you have a couple of other extra nodes, and the Services that we're running on. There are info blocks, ddi, or dhcp, dns, and ipam is Running under controller with open stack. And paul walter network's v-series firewall and the Big five, rather f5, big ip, ltm, traffic manager is running on One of the compute nodes. So that's one of the setups. The second setup is similar. This is running on canonical open stack. Again, the jr release. But what's differentiated here is that there is a metal as A service server that rolls out additional servers as Load requires it and scale is needed. So that's one of the ubuntu value adds. But the rest of the setup is very similar. Where you have the gateway for talking with the underlay Network and the obby network controller and the service Engine running on a compute node. So having done this, I want to point out something about our Setup. So we have a single network controller for both of these Open stack reference architectures that are currently deployed. It's one network controller with two open stack distributions Underneath and they're separate from each other. One is based on the red hat distribution and one is based On canonical. So it's a multi-tenant environment that could be used Within the enterprise for different environments for Different use cases or within the service provider based On the customer preference. So customer A may want red hat or they may want open Ubuntu, we have that capability. But just as easily, because we announced a partnership With moranthus today, we could extend it with a third or A fourth or a fifth instance of open stack scaling this Out. But again with a single SDN controller in a mix. Now the SDN controller itself has a peering capability Where it scales out depending on the need of the data Center. But from an operational simplicity perspective, it's One network that manages all these separate cloud Environments simultaneously. So in that sense we're sort of like the one ring that Rules them all. We control it. Except the difference is this guy is reaching for the Ring. We have the ring in our hands. We're the controller that controls them all. So let's throw that away and let's talk about the Physical infrastructure. These reference demos were deployed on soft layer. In four different data centers across the world. San Jose, Toronto, D.C. and Hong Kong. What we did was overlay a logical land across these Four geographically separated areas. So from an application perspective, whether your Workload is running in Toronto or in Hong Kong, it appears To be on the same land. Now you saw before that I talked about six different Servers. The six different servers are distributed in these four Different data centers. But to the application it appears that everything's Co-located together. So that's a way of reducing complexity while increasing The redundancy of the network and the resilience of the Underlying structure. Because if you have a disaster in one of these cities and The data center goes down, you have the capability of Bringing up similar services and from the Applications perspective it doesn't matter where the new Data center is. The network again makes it seem like everything is on The same single network. Now next I have a very brief video here that will not do A justice to the demo because it's actually, it's Involved and there's a lot of fine details. I will play this video but what I would encourage you to do Is to go to any of the partner booths, ours, or There we go. Ours which is back there, red hats, info blocks, Canonicles, we all have these demo running that shows how The setup injects services into an infrastructure setup, how You can deploy networks, and how the policies Propagate from let's say polo alt and firewall down to Vsd and can be managed together. So please stop by our booths because given the 20 Minutes here and the resolution on the screen that was Worried that by giving you too much detail it would Lose it but you can quite easily see the same demo in our Booth and the various partner booths. So with that I want to thank you for your time and Invite you to come to our booth and know our networks. My name is Amir Sharif and I'm available for questions After this demo as needed. Thank you very much for your time.