 Thanks, Shannon. I want to take a moment to introduce Aconda, which is OpenStack integrated layer three through seven services. A little bit about me. My name is Mark McClain. I am the CTO and co-founder of Aconda. A member of the OpenStack technical committee was one of the original lead developers on the Aconda projects. The Aconda project originally started back at Dream Hosts if you were listening to Jonathan's talk prior back in 2012 and so it grew out and then late last year the Dream Hosts team spun it out into a separate company and so that was how Aconda equips born. But one of the things that started originally with Aconda is we had several goals in mind. One of those was to deliver a very simple and elegant solution that was easy without the need for complex SDN controllers. Make it compatible with multiple layer two fabrics. What we wanted to do is enable deploy your choice. That way you're not having to install a very monolithic blob SDN controller within your deployment. You have options for which layer two you're running. You have options as well. Within Aconda we'll talk a little about that as well as Aconda's open source. Like I said earlier, Aconda was born in Dream Hosts. Dream Hosts has a very strong open source core to the company and so Aconda has been Apache 2 licensed since day one. In our current release, which is Aconda 1.0, we announced it just prior to the summit. It includes a network, it includes a orchestration platform which we refer to the rug, which is kind of a reference to the Big Lebowski, which was the original code name for the project. Aconda supports multi-tenant network services including routing and load balancing as well as support for third-party network services. The reason we included support for third-party network services, while we at Aconda will integrate and give you, say, our version of a router or an HA proxy-based load balancer, you have the ability to bring your own software router that can be integrated through our driver or as well as like more recently this week we announced a partnership with Nginx Plus, which integrates and has a driver in Aconda as well. The 1.0 version supports multiple overlays, so Aconda will run on top of the Linux bridge, it will run on top of OVS, it will run top of OVN, it will run if you're using hierarchical port binding and combined Linux bridge on the hypervisors with cumulus switching at the top doing the XLAN offloading. So you can deploy it a number of different ways, we've tested it in those ways. Aconda interfaces using standard open-stack APIs for Neutron Nova Solometer and Horizon. One of the questions that comes up frequently is, are we doing anything special with Neutron? And the answer is no, we're using a stock Neutron and then have drivers that meet with the traditional layer 3 plug-in interface for Neutron, so no secret or special sauce. So if we take a little look a little bit of the architecture, what you're going to find is you're going to find your physical layer 2 network underlay. If you're running overlays with the network, Aconda has agnostic overlay support, so as I touched on a little bit earlier, from open-source versions you can have something that's based on primarily OVS or Linux bridge, you have those two choices. We've also tested on a few proprietary overlays and it works. And then between that, the L2, if you're running overlays are typically managed by Neutron and ML2 or plug-in, and then you have the traditional open-stack APIs and then on top of that, Aconda provides managed services for routing, low balancing, and then coming with support for firewall as a service. And the Aconda management sits alongside of all these. So if we kind of consider what Neutron looks like without Aconda, this is kind of the traditional diagram you'll see. You'll see multiple layer 2 agents, you'll see a layer 3 agent and DCP agents and advanced services agents. And what occurs in those deployments is the L3 agent, DCP agents end up being choke points and single point of failure. Yes, you can have multiple copies of those services running, but even with L3 agents, if you're not running in DVR mode, you end up creating a choke point for really congestion and within your deployment. So with Aconda, we can actually simplify this story significantly because we can replace the layer 3 agent, the DCP agent and the advanced services agent with a single Aconda service which manages and orchestrates the network services for you. So now time for a demo. So what a first one to point out is if you look in this window on the right hand side of the screen, what you'll notice is I'm just tailing the log in the background. This is just the Aconda Orchestrator service. You'll notice that it's taking actions and states. For deployment topology, I'm just using a very basic install of OpenStack with Horizon. If you notice, we've got a router here. We've got a service, we've got a VM up. And the VMs are just running CROS configured to use DHCP. And so if I go in and create a new network, so we go through and create a network. If you notice in the background, the Aconda Orchestrator is working, it's received the event, it's noticed it and currently because the network's not plugged into any routers, if you notice when our topology map, it's not connected. So if we go in to the router and click add interface, I can go in, I can select our other network. And if you notice now, and if I go back to our topology view, you'll see that the Aconda is working in the background. We now have our two tenant networks. So if I want to just kind of show you a real basic demo of just show you how Aconda has wired both networks in for routing, I'm going to know I'm now going to create a new instance on the other network. And if we go back to our topology view again, you'll see that the instance is up and running. If I click on open console, if you notice it's art, the DHCP was automatically configured in the background by listening to the streams. I can ping the other instance. So in this case, what it's doing is it's routing between the two, the two tenants network segments. And so this is Aconda working in the background. Now as a deployer, one of the other things that's kind of useful is Aconda has a full CLI command line tools, but also sometimes we've integrated Aconda in with Horizon. And so like I said, our orchestrator is called nicknamed the rug. So if you go down here and click on the rug tag, what you'll see is you'll see a list of the routers which Aconda is a managed. And so in this case is set up for routing. So it's listed the routers that it's managing. The instance name it's running when it was booted. And also you have a couple of actions you can take. So for those who need to go in and fix one or two or kind of investigate, you have rebuild router so that you can cause the router, the orchestrator rebuild it. You can place the router in managed mode or unmanaged mode. So in unmanaged mode, it gives you the ability to say if you're in a testing environment and you want to investigate or you want to look at different configuration options, you can actually make configuration changes because one of the things that Aconda is watching for and needs to hood is it also watches for config drift. So if you go through and make manual changes and your routers in managed mode, Aconda will go back and reset the configuration state to the proper logical config. But sometimes when you're working in an environment operations mode, you don't want Aconda to replace your handcrafted. So you have the ability to place it in debug mode or in manage mode or just force a config push onto the router. As well as you can set whole tenants into managed mode or unmanaged mode. This is useful if you have a test tenant and you can use that to your advantage. Typically what we'll find with the players is when they're rolling out new versions of the appliance image, they'll have a special tenant they use for testing via the rebuild command. You can go through select which image you want to boot with and then it will use that to boot with the test image and that way you can make sure for any compliance testing. So one of the other things is so if we're taking a look at the Aconda in the background is also looking for changes. It's constantly watching and keeping the health. If you've noticed when I've been talking in the background you'll occasionally see some text scroll by. So what Aconda does is it constantly goes through and checks and via the driver interface we're able to have a pluggable way to detect whether the appliance is healthy or not. And so by doing that it's constantly monitoring and so if you were to say kill an instance or a hypervisor were to die, Aconda will notice that and check and we can spin up a new version. So what actually is an Aconda? Routing, Aconda supports IPv6, supports standard DHCP and metadata. Aconda also supports BGP, internal BGP as well as one of the things about Aconda though is wanted to be transparent to the user. So as you notice from the open stack interfaces you couldn't really tell we were using Aconda. The tenant interfaces looked as you would normally expect. All the logical components were the same. There wasn't anything it called out because mainly Aconda is mainly for the back end for the deployer side. Also as I said earlier it's delta agnostic so whether you're running OVS in your hypervisor or Linux bridge in your hypervisor it will work there. And so this is kind of the topology for the slides posting later for where we added one in. As far as and I mentioned the rugs with this auto failure detection and showed you pictures of the CLI and the horizon support which is in beta. And then kind of just kind of give you overview of under the hood how all the components fit together. If you were say have L2 agent running Linux bridge you could have those could be talking VLANs to the top of rack and then from there the rugs actually listening into a special management network that's connected via the overlay in so that way the rug communicates with the router. So if you see the two orange boxes are communicating with each other and so here you have a hardware VTAP you can have a chemo switch running as a VTAP. For example horizon integration so just kind of the recap Akonda is 100% open source project it's been in it's been developed dream host since 2012 it's used by thousands of VMs daily it's in production so don't be scared off by one point oh this is actually a production system. We were spun off like I said last fall is a separate entity and so our one dot oh is available now. So the key features just kind of recap is it's 100 compatible kilo it runs with a vanilla kilo installation basically routing basic load balancing service management layer 2 agnostic down the road we we're going to add some we're going to add extra features as we go through the summer so join us we have community like I said it's all available on it's all open source so GitHub slash Akonda we have IRC channels if you want more information on PSEs or any kind of testing info at akonda.io or you can come find us in our booth over there in t50 so thank you very much