 I don't think anyone is new in the room, but I guess I'll reintroduce Masayuki again. So this is the Cloud Containers and DevOps Track. Masayuki Higawa is a senior software engineer at SUSE, and an open stack developer, and he's going to be talking about free and open source Kubernetes management tools. So without further ado, Masayuki. OK, thanks. Oh. OK. So I will talk about, sorry, I'm Masayuki Higawa, and I will talk about free open source software Kubernetes management tools. So this slide is on here. And so my GitHub Twitter free node is Masayuki G. So if you have any questions or contacts, please ping me on that. Yes. And I think almost the people are the same, right? So I am an open stack upstream or downstream programmer. And my job is a quality assurance in the upstream or the downstream. I mean, I'm working at SUSE. I'm also a quality assurance member of the SUSE OpenStack Cloud. And I'm also a methods user group in Tokyo. So this is a bit update profile. And in this talk, I'd like to talk about Kubernetes. So why do I like to talk about Kubernetes? Here's the reason. So first one is it's a famous and interesting topic, technology. So today or previous days, many people are talking about Kubernetes or Docker or like that. So I also, I'd like to know the technology. So I tested some in my cloud. So it's a free open source software. And I also use it in my cloud. I mean, not Google Cloud or Amazon Cloud or Azure. So I'd like to talk about my private cloud. It's three one-use servers plus one workstations. It's powered by the Lack Rock. It's already talked about Matthew in this morning from 10 AM. Yeah, already talked about that. It's very similar. The difference is servers spec and the number of servers. So if you want to see more, if you would like to more, you can see the GitHub page of Matthews. And the difference is also Linux distribution and the OpenStack distribution. I used the OpenStack Queens version on OpenSuzerLip42.3 version. And I set up it from OpenStack distribution. I mean, just use a package, not a turbo or a Git source. So it's really simple. And I have no trouble with that, basically. But there are some issues on that. It's a network. I build it in a very flat network. VMs are connected to my home router. So it's a very simple network. I can show you my network environment. Well, let me. Yeah, like this. So VMs are connected to my home router directly. And my workstation also connected to the router. So it's very flat. So I don't need to split the network. I don't want to split the network because it's a bit complicated to me. So here's what is Kubernetes. Maybe you already know that. But it's a container orchestration tools. It has a lot of features like self-filling, automatic binary packaging, horizontal scaling, auto-scaling, load balance, something like that. And it's open source. But I've heard that it's a bit difficult to deploy it, especially in a production environment. So I like to use it easily. So I checked some first tools. So why we should use the first tools? It's free, not only money, but also you can see the source code. And so we can know everything. It should be fun, right? And I have the resource. I mean, I have my cloud. I can run on that. And so first, Kubernetes tools. There are some tools like this. MiniKube is kind of a, I think this is not only for, this is not for a production environment. It's only for developers' test environment. And the second one is OpenStack Magnum. It's one of the OpenStack projects. I'll talk about that later. And the third one is Metos, DC OS, and Launcher. So I'd like to talk a bit about Magnum. It's an OpenStack project. And it's API services for container orchestration engines. So this Magnum not only supports Kubernetes, but also supports Docker's form or Apache Metos. And I only read the document. And it will support DC OS too, but not yet, probably. And Metos DC OS is a, Metos and DC OS is a kind of different. But DC OS is based on Metos, right? So Metos is an abstract CPU memory storage and the computer resources. This is similar like OpenStack. But it's also enabling the Fortoriant and Elastic distributed system to easily build and run effectively. Yeah. And DC OS is called Data Center OS. It's an open source distributed operating system. Yeah, it has more features compared to the Metos. And I think DC OS is one of the point is it has fancy catalogs. So probably I can show you if I have time later. And you can see the information here. And then Rancher is an open source too. And it's a container management platform and the deploy and manage the Kubernetes with ease. So I tried to use it. And I found that it's very easy to use it as a first step, as a starting point. And this also has a fancy catalog. So it's really good, I think. So there is a plus cons. I tried to use the Magnum and the DC OS and the Rancher. So here is a plus cons. So first, I tried to use the Magnum. But I failed to build the environment. There are some reasons. One reason is Magnum requires a separated network. It doesn't fit my open-source environment. So it's one of the reasons. So if I can use the Magnum, it should be used with the standard APIs, I mean the OpenStack API. So if you are very familiar with the OpenStack API, it should be a good option. And the second is the methods, DC OS. I need to change the sentences. So methods needs several nodes at least. It's one of the issues, not the issue, bad point. But a good point is it already has a full tolerant. So it should be a good point, too. And Rancher, it can be started from one node. It's really good to use. So people doesn't have a lot of nodes, basically. So it's really good. And Kubernetes CLI included. But I think CLI4 to manage the Rancher is poor or not. But there are some APIs. So we can use that from that API if we want to use it from the CLI. So for future work, this future work is just for my project. I mean for my private cloud. So I need a real use case to use like these Kubernetes management tools. So one candidate is IoT things. It should require a lot of resources or like that. And as in test environments, it's also good use case for Kubernetes. And I think I should have more stable infrastructure. But it's not so unstable for my opens to 42.3 right now. So it's not so high priority. And the third one is introduce network isolation to use OpenStack Magnum or like standard OpenStack Cloud. And the last one is, yeah, I'd like to use Magnum. So here's the conclusion. We can play the Kubernetes with the posts tools easier. And having a private cloud is very fun. So I'd like to show you kind of a demo. If I still connected to my cloud. So here's my cloud dashboard. So it's not so different from the standard OpenStack Cloud. So you can see already I have 1, 2, 3, 4, 5, 6, 7, 8. VMs on my cloud. And there's some variety of the VMs. I run the OpenSuser, Debian, Ranchoist, and the CentOS like that. So if we have the private cloud, we can run a lot of VMs like this. And I'd like to show you the DCOS dashboard. It should be work. This is work on my cloud. But there are some network issues probably. Yeah, so it doesn't work. But yeah, if it works, we can see that a lot of application catalog on that page. And just quick the applications, it will be deployed to the DCOS methods cloud. So it's really easy to use that. So I think this doesn't work actually. So this is the end of the week. It's how many hours? OK. So that's all for me. Thanks. So any questions in this part or the previous CLI part? Whenever, whichever you want. OK. Thank you. Thank you very much.