 Good afternoon, ladies and gentlemen. Welcome to our demo. Today, I will give you a live demo about auto scaling over OpenStack hybrid cloud. First of all, I will give you a short introduction to our company. We are EasyStack, which is the name of our company and also the name of our product. We are the leading OpenStack service provider for China and Asia. In June of the release, we have contributed a lot. And we are the number two contributor from China. And we have been constantly contributing to the community until now and for the future. Right now, we are focusing on OpenStack services and solutions. We provide various solutions to our customers and consulting and product deployment services to our customers. This is an overview of our solutions. EasyStack, the us platform, which is based on OpenStack, we have software defined computing, networking, and storage. And on top of that, we have billing solutions, monitoring, and a ruler, which is a deployment tool to quickly deploy our solutions and the product to our customers. The EasyStack connector, which is my focus today, is a connector between our private OpenStack network to the public cloud. It provides the secure VPN channel between OpenStack network and the public cloud. So for today's demo, the hybrid cloud auto scaling, our solution uses the following key technologies. First of all, it's a VPC-based overlay. It provides a secure channel between OpenStack network and public cloud network, your own tenant. Also, we have a distributed virtual router, which reduces bottleneck of a neutron L3 router. The compute node can handle the north-south flow, so we don't have the bottleneck. IPR router, which is also critical because you don't have a lot of public IPs. If you can allocate the IP address to your router, then you can do some port forwarding from your router to your subnet, right? The next one is a hybrid connector. This is the core of our solution. It can dynamically create the secure VPN channel between your OpenStack tenant and your VPC or your tenant network in the public cloud. Our auto scaling in this demo is best on hit, but you can also use other solutions to work with our hybrid connector. Also, from our dashboard, you can see the layout of both your private network and your network in the public cloud. Now, let's see the architecture of this demo. This is a very simple scenario for this demo, but the solution also, you can have other scenarios supported on our solution. In this demo, I have three kinds of virtual machines. The HAProxy for load balancing, the flow, the traffic, or the WordPress. The WordPress uses MySQLDB on the DB virtual machine. At the first, all the VMs are in our private OpenStack cloud. When the load become high, we can auto scale the WordPress VM to the public cloud tenant. So the HAProxy will load the balance between the WordPress in the public cloud and also in our OpenStack cloud. The WordPress in the public cloud also uses MySQLDB in private cloud. You can do the same for the DB clustering, such as auto scaling the DB VMs also to the public cloud. The easy stack connector here provides a secure channel between the public cloud and the OpenStack private cloud VPC. And it's very secure because from the public network, you can access the WordPress VMs. With the introduction of this architecture, I will show you the live demo. This is our dashboard for easy stack. And to enable the hybrid capabilities for each... Excuse me, OK? Please give me a second. I'm sorry for the issues. To enable the hybrid cloud for each tenant, the admin have to manually do this because when you enable the hybrid, easy stack connector will allocate some resources to set up the channel for the tenant network. And sometimes you may not want all the tenants have the capabilities of hybrid cloud. So we have these settings here. After you do this for the project, the project admin can set up his own public cloud accounts and credentials from his own tenant here. I take the AWS public cloud for the example and to save some time, I already set up the account. So in this project, we can freely scale our workloads to AWS. Also, to save some time, I already use Hit to create the stack of our workloads. Three VMs now for the HAProxy, WordPress and DB. Also, I set up the internet access here. You can see that I can access my WordPress application here. Right now, for instance, if we have high load of the WordPress from users, then how can we scale the WordPress VMs to the AWS cloud? We use the Hit alarm to do this, to trigger the actions, to scale the workloads. But I don't have the load now, right? There isn't many people accessing my WordPress now, but I have my colleague who has a simulator to increase the load to my WordPress and trigger the alarm, which will scale my workload. Right now, I ask my colleague to do this. He will notify me when the workload is active, and we can see we can go back to the network topology. As I mentioned, you can have a unified layout of both the private cloud and the private cloud network and your tenant network in the public cloud. We have slow network connections here. OK, at the same time, waiting for this page to load, we can go to the WordPress and see this again. Refresh this page. It's also working. But you don't know if my AWS VM is working, right? To prove this, then we can shut down my previous WordPress in the OpenStack cloud and leave only the VMs in the AWS and go back to this WordPress to see if it will work. Right now, I will shut down my WordPress in OpenStack. It's loading. Anyway, I think it's shut down now, and we go back to WordPress and refresh this page. If you may know that the network connection in China to access AWS is very slow, so it may take a minute to load this page. OK, this page is completely loaded. And right now, only the WordPress in AWS is active, so I already show you that our scaling is working. It's too slow here, so OK, it shows up. You can see the AWS VMs from here. It's triggered by my colleague. And right now, this WordPress in my OpenStack cloud is dead. Only these VMs is working. Also, you can see the networks here. The first one is a private network in OpenStack. The third one is also the private network in OpenStack. I can think that this tenant could connect both of these private networks to AWS with secure VPN channel. The second network is the network from AWS. And all these AWS VMs are connected to these networks. So you have a unified view to your VPCs, both in private and public cloud. Also notice this router. This router is for the tenant. And you can create freely and allocate a floating IP to this router. We also have the scheduling mechanism to schedule the router to computer node. And it has active standby, high availability. If the router is dead on the computer, on the computer node, the router will be rescheduled. The standby router will take effect. OK, I think this is all the demo for the hybrid cloud. If you have some questions, I have 50 seconds. I can take one question. OK, if you want to know more about our solutions, you can go to our booth right behind this theater at E15. Thank you.