 So thanks, everyone, for joining me. Basically, we're going to go into an overview of SUSE Cloud, which is based on OpenStack Havana and our high availability portion. This is our deployment framework so that you can easily deploy the components of OpenStack. We've incorporated a pacemaker and a cursing into the equation so that we can have a highly available controller node. So this is not your live migration between compute nodes. This is for your controller node. We start by creating a pacemaker cluster. And we can choose the method for an SBD device or different mechanisms to configure the cluster for internal communication. In this scenario, we are using an SBD device with Dev SDC, which is our split-brain device for the split-brain detection. And we're enabling DRDV for certain components, like a RabbitMQ or MySQL. And we're also making available the all-style HB GUI interface for pacemaker. We can then easily choose our known servers, which in this case, we have a two-node cluster. And then we pretty much apply the proposal to deploy our pacemaker cluster. Moving forward, the database, in this scenario, we selected a high availability using DRDV. We're using one gigabyte as a storage for now because it's just a demo. Normally, depending on your environment, you may want to have 10, 8, 15, or 20 gigabytes here. The end result as we start creating these components into pacemaker is what we show here in Hark. We can see all the resources for neutron, glands, RabbitMQ, Keystone, pretty much every single resource out there. What's missing here to deploy is Horizon. So we will have a session Thursday afternoon, hands-on session on how to do this from zero to full deployments. So if you want to join that session, we will gladly have you there. So part of the value here, of course, of having a high availability is you can do a migration of the resource. If you want to migrate, let's say, from upgrade from Havana to Icehouse, so you can have one node and be upgraded on a different version of OpenStack and then failover again to that node. So there are a lot of use cases, and of course, eliminating the single point of failure, which is the main goal here. And let me see what else I can show you here. We can create a Horizon proposal just a moment. I'll get back to Keystone, actually. So this is part of the configuration for Keystone. And we can configure the default values, assign the components to the cluster we created with Pacemaker. And our developers have made pretty much the process so that both components will get installed in each node of the cluster, yet the resource is obviously in one of the nodes. It would be the active node. We can see that in Hock specifically. So this is pretty much what we wanted to show today. For a full hands-on demo, please visit us on Thursday afternoon. You will need some files. You can get them at our booth from tomorrow on. And if you have any questions, we're right in the booth, and I'll be around the area.