 All right. Good afternoon, everybody. My name is Hussain Kazal and I'm here with Remy Vishary. We're from Nuage Networks and we're part of the technical integration of our partners within Nuage Networks. So today we'll be talking about the topic is one small step for SDN, one huge leap for Dev app community, and that's basically we're going to talk about what Nuage does, give you a little bit overview for the people who are not familiar with Nuage and then one of the challenges we faced when we were building our ecosystem and how we were able to overcome it using OpenStack and Nuage and we're basically launching a new service using that product. So for the people that don't know we're Nuage Networks, we are based in Silicon Valley. We're a little over three years. Anokei Aventure focused on leveraging our expertise in networking and bringing that to cloud environments. To sum it up, we basically connect anything, any workload, anywhere over any infrastructure. It's a completely software-based virtual solution that spans multiple data centers, and we've been a member of the OpenStack community for a while now. Architecturally the product is three tiers. At the management layer we have the virtualized service directory, which is basically our policy manager. It has a northbound API interface and basically manages the entire overlay network. In the second tier, it's a controller, which is based on Alcatelus and Nokia's service router where we virtualize the control plane and that enables us to scale. So as you're at the number of hosts that you have in this are increases, you basically add more controllers and they peer. The advantage here is actually you can peer with another PE router and that provides you the multi-data center capability. On the data path, we have the VRS, virtualized router and switch, which is OVS based with modifications in the user space to allow us to provide distributed routing. The advantage there is that we said we connect any workloads, so containers, virtual machines, bare metal servers through a gateway or by installing our VRS on the bare metal itself, the VRS-B. And then the gateway is we have a soft gateway that will provide VLAN to VXLAN connectivity, which is our overlay solution or basically using a physical gateway where you plug in whether it's ours or a partner such as Arista, HP, Dell, Dell Cumulus. And when we look at least our focus area is building a technology partner program around the product. As a networking product we basically have multiple touch points with the partner solutions from management and orchestration such as service orchestrators, cloud management systems such as OpenStack, CloudStack. From appliance perspective, we service chain, so whether you're a security service or appliance or a load balancer, whether you're a platform such as a, like you have a hypervisor or networking products such as a top rack switch and then we have the system integrators. So looking at all of those things we and you know, we start with a small number obviously of partners and then we over the years we were adding those partners and the thing that enabled us to do that is that there's nothing really custom about our integration with the partners. Everything we do from a networking perspective is API driven, so all of the integrations are standard. So when we work with an appliance such as a load balancer, we're using LBAS, what a V1 and V2 and what we do with one partner, we do with everybody. So that enables our customers to kind of build the best of breed solutions. As you can see from the footprint, it's very difficult to, with four people, to be able to do all of these partnerships and most of them are technical integrations where you need the partner to kind of deploy your platform, learn how to use it and then develop code to communicate with it, whether it's to provision or pull information out. So to formalize our kind of partnership program, we're launching a sort of a technology alliance program. We call it the Nuage VIP program and that basically helps us to build a three-tier model where depending on the level of engagement that the partner wants from us in terms of strategic preferred or access, the more engagement that they want to do, it's obviously a strategic level and then if they want to self-serve, it's the access tier and that's around mostly development, but also includes support sales and marketing for the higher tiers. If anybody wants to, if you have a product or you're a developer that wants to build something on top, launching the program, you can go to our website or email us to kind of join it. We'll provide you with a guide and all the benefits that you get. One of the tools that will enable you as a partner to build an integration with Nuage, we're calling it the Nuage Networks Experience, which is basically a service that we're launching this week. And as mentioned, you know, we consider the platform very dynamic, very flexible, very scalable, and you know, all of the cool features that we have are great. We open sourced our VSPK, which is an SDK a year ago, it's a wrap around our API, so there's documentation on the links I have. You can go and basically download it and then you'll be able to use it to develop an integration. The challenge there was in order for you to take advantage of it, you actually have to have a Nuage instance installed on-prem, so you have to record resources, install it, and that was a challenge that we're overcoming today with Nuage X. So what Nuage X is basically targeting developers, customers, and partners to enable them to really leverage that VSPK and do an integration without having to install anything on-prem, right? And the main use cases are testing curiosity, if you just want to figure out what Nuage is about. If you're a partner that wants to build an integration with Nuage to kind of insert your service or to pull information, basically all of those, if you're a customer that wants to test our next release, test your APIs against a future release, this is kind of another use case for this platform. And the way it works is basically we are hosting in our data center a bunch of servers and then we'll allow you to log in, deploy our management and control planes, and then connect any host that you own, whether it's in a public cloud, in a private data center, on some of our partners, you basically, as long as it has IP connectivity to our public instance, you'll be able to connect all of your workloads to that control management plan and be able to spin up containers, virtual machines, and then test whatever you want, develop whatever you want. And that's kind of also one of our partner, like if you're a partner that wants to run a demo with Nuage, you can basically, you don't have to install the management control, it's basically we will spin up. And the process is very simple, you go to nuagex.io, you log in, you basically select from a list of available labs, depending on the use case, and then once you deploy it will actually, depending on the environment that you're using, if you're a CentOS shop, you went to, we will auto generate a script that will automatically install our VRS on your hypervisor connected to your lab. And then the next thing is basically you just run containers or spin up VMs and they'll be controlled using our control and management plane. And all this, the key thing here is how quickly you are. So the aim is like less than five minutes, you're logged in, deployed the lab and are able to connect your host and spin up VMs and containers. The website is nuagex.io and then it'll be launching, Friday you'll have basically the website to create an account and then get credentials. And Remy is gonna, to the interesting part, which is the actual demo and he'll talk about how we built it while the demo is set up, sorry. Thanks for saying. So for those of you who are not there at the beginning of the session, my name is Remy Vichori, I'm an intact marketing expert at Nuage Networks. I'm mainly focusing on technical integration with partners, but today, as Susan mentioned, we'll not talk about integrating the partners. What we'll do is actually a real cool live demo about leveraging the Nuage platform by deploying like thousands of containers in a matter of minutes using a Nuage Networks experience platform and some Bermittal servers deployed on packet.net. Actually we requested like some Bermittal server from packet.net which is like a public cloud provider for Bermittal assets and we deployed a lab on our premise to be able to run this demo. So let's jump into the actual architecture. So what we did beforehand is we deployed VSC and VSD using Nuage Networks experience and we have requested 25 server from packet.net across two different data centers. One is hosted in Newark, New Jersey and the other one in Sunnyvale, California. So actually we'll deploy thousands of containers, so maybe 10,000 across completely two different data centers. And as Nuage Networks, we love challenges. We'll not do 10,000 on 25, but we'll actually do 25,000 and we'll try to keep it before the end of the timer. So it's actually demo time, but just before jumping into the deploying the containers, I will just like do a quick walkthrough about the Nuage Explorer. So basically when you'll have an account and register, we'll activate your account and you'll be able to go to the screen and actually use your login and password to access the dashboard. Obviously in this dashboard, we already have a lab already deployed because that's the lab that we'll be using for the live demo. But the steps to deploy a new lab will be really simple. We want you to be able really to deploy new HBSP in a minute. So what is by doing just like clicking on the plus button on the bottom right corner, you just fill up a name for example, OpenStack Barcelona, and select actually a date where the lab will expire. And you'll click on it and it will start, normally it will start, oh, I reached my quota actually. So there is a quota one-on-one. So I reached my quota, so I can deploy it. But if I don't know that I can't, I will show you. It will start deploying in a matter of minutes. The lab will be ready and you'll be able to see the same thing of that. So the lab will be started and you'll be able to access the environment. So what will happen there is you'll have the credential to access your dedicated instance of VSD. And by clicking on Access VSD, it will open a new tab and start like getting the VSD architect page. Let's jump to this one. So, yeah, it's getting there. All right, so I'm looking to this VSD instance that is deployed only for you. And I see that I have actually zero containers, but we have deployed some subnets beforehand because obviously we need a topology before launching all these containers. So now let's switch to the most interesting part and the CLI commands. So I separate the screen in two parts. On the left side, we'll have a counter and a timer that will increase as soon as we deploy more and more containers. And on the right side, I will just launch a script that will connect to the 25-horse and deploy the containers. So as you can see, the containers are starting to be deployed. And what you see currently is the number of Docker containers really running on the VSP platform. So what happens is we don't take the counter from the Docker host, but we take them directly from our management plane. So that means that when the container appears there, it means that the IP address gets already allocated and all the security policies are already distributed across the 25 hosts. So as it will take like maybe, I don't know, five or six minutes if we can't get there, I will just talk briefly about how we use OpenStack and why we use it to build this Nuagex platform. So when you're a go, we had this idea about creating a demo platform for our partners. And we started to build some service on it based on OpenStack Juno and the Nuage network VSP. And some weeks after releasing it, like internally, we saw that it was a great idea to share it with our customer, not customer, but employee to be able to leverage that and deploy the demand in front of customers. One year later, we have almost deployed, I think, 14,000 VMs on this platform. So and we have like a really, really good demand, maybe around 300 users or something like that internally. So this was nuts, and there is no joke there. It was really cold nuts. And now it's Nuagex. So what we did is we took the nuts folder and split it in multiple microservices to be able to quickly add new features. We had like so many features for the Nuagex, like organization, documentation, templates. There is a bunch of new things that are coming with the Nuagex release. And we have taken the past to upgrade also to OpenStack Liberty to leverage new functionalities that are available in it. For example, it's Stack Snapshot because you'll see some of our employees want to snapshot their stack and be able to restore it before going to a customer demo. So one of the challenges was to, we were deploying the assignments, but we had to link those environments to a rental platform which is running like Docker containers, actually, or our management plane is done in containers. So every service that you saw in the previous slide is deployed as one or multiple containers. And what happened when you click on deploy? Actually, it deploys a new lab on OpenStack. And also it deployed a new container on the rental platform to actually monitor the lab. So what was the challenge there is we needed to put that container in the same network as the VSC and VSD. So by using the Nuage network VSP and Delay, we were able to seamlessly connect both the Docker containers and the VM together. So now let's see if time's good for us. Let's switch back to the contours. So actually, we're already at 12,000 containers. So maybe I talk too fast. Yeah, you can go back to the last slide and just scroll all the way down. This one? Yeah, just this one. And then just talk about multi-tenancy maybe. Actually, what we did also there is we tried to leverage our VSP platform really and use all the features that are available. And one which was really interesting there is as a end user, you'll be able to access those VSD and VS instance. But we didn't want, for example, one customer or one end user to be able to jump from this host to the other one. And we used for that like Nuage Network's micro segmentation to tell that anything that is going from AMP to any other environments will be blocked. And we only allow environments to get out of the network or come in on a special port. So by doing that, we are securing well our environments and we are sure that nobody can come in on a lab, mess up the lab and jump onto other things. So we are sure that we are not propagating like malware or things between different environments. So it's almost one minute before the end of the talk. You can go to the last one and then we can go back to the demo. Yeah. Okay. So just to conclude this talk and then we'll jump back to the demo screen to see where it goes. Nuage Network is really an open platform because we have open REST API. All the REST API documentation is available on GitHub, the API specs, the latest API specs are also available. We have released a couple of SDK in Python, Go Java, and we are soon to release like a new SDK not for our REST API but for actually managing the data plane. As you said mentioned, we just launched the Nuage Network VIP program so it lets our partner to have a compelling solution together. And by Friday you'll be able to access the Nuagex.io website and click on the registration and we'll start to activate accounts maybe every week as beta users. And let's see, if you have any question or want something like to bring back to your family, just came by Boost C32 and we're really happy to answer any question. Sure. Let's come back here. So here we go. So we had deployed 20,000 containers in actually six minutes and 23 seconds. So that was really cool. Thank God, I think the coolest thing about that is I think maybe other vendors can do the same thing but what is really powerful is we have pushed not only IP addresses but also we have pushed policies across the old cluster. So what we'll do next is we'll just get all the IP addresses and all the names of all the containers, so the 20,000 containers and we'll start selecting randomly some containers just to learn something between the endpoints. So what happened is the container 10, 15, 55, 90 as a source will ping like container 0, 1, 2, and 3 and you can see that actually the ping are working. And the container can be deployed like in New Jersey or Sunnyvale, we actually don't care. For us it's actually the same network. Thanks for our attention. Thank you. Thank you.