Welcome to this lesson, Deployment Models, Hosted Solution Options for our course, Deployment Kubernetes. Kuberneter has a number of different installation configuration options. However, it is the same Kubernetes just scaled up in different ways that go from a smaller project level type of installation up to a full enterprise distributed solution. One option here, the single-node installation, very, very similar if not identical to Minikube, is where all of the Kubernetes cluster is installed on a single host node. The cluster or the Kubernetes cluster is virtualized in that environment and the master node and the worker nodes are all hosted within that virtualized single environment on that piece of hardware. The single-master node with etcd and multi-worker node installation is where the master node is installed with the etcd on a single-host node. Then connections are made to work to multiple worker nodes either on virtual hardware or on physical hardware. And with these options, this is where we begin to scale up into full enterprise distributed type of solutions. We have single node etcd and multi-master and multi-worker node. So, the multiple-master node configuration is in high availability, but there's a single instance of the etcd on its own host. And here, there are multiple worker nodes connecting back to the primary master node. And then, the most distributed solution is where we have the multi node etcd with a primary host. And then, there is a multi-master node and multi-worker node installation. So we have multiple etcd and masternodes all cluster together in high availability mode. So there is currently running one primary etcd and one primary master. However, if any of these hosts go offline, since they're in high availability mode, another host will step up and take its place. Connections are then made to multiple-worker nodes. This is again, the preferred enterprise production configuration. The single-node installation is the simplest Kubernetes configuration. Single node is rarely used for production use cases. Single node is useful for non-production environments or for sandboxes or for training. Single-master node with the etcd and multi-worker installation is suitable for production environments. It's scaled up. It's more scaled up than the single-node installation. This configuration supports pods running on separate nodes and provide redundancy in the event something goes wrong with the node. We can see here, we have some variation of our master node connecting to our worker nodes. And this is always the same Kubernetes architecture. It is just how it is installed on the various hosts. The single-master node with etcd and multiple worker node installations is also suitable for production environments. This configuration supports pods running in separate nodes and provides redundancy in the event something goes wrong with a node. So we can see here again, we have our master node and we have multiple-worker nodes. This configuration here is the multiple-master node. So this is high availability mode. There's a single etcd keystore running on its own host. And then again, we have multiple-worker nodes. So you can see here that there is high availability in the event something goes wrong with our master node. And we also have multiple worker nodes that also provide redundancy. And here, we have the multi node etcd, the multi-master node and the multi-worker node. So we can see here that we have our high availability etcd on their own respective hosts and high availability. Again, we have our master nodes installed on multiple hosts thereby supporting high availability and then we have our multiple-worker nodes. So this is obviously the most distributed, most flexible, most enterprise worthy type of installation. It is for absolutely mission-critical installations, but along with all of those benefits, come some levels of complexity. Again, though as you can see, if you look closely, it is all the same Kubernetes, all the same architecture. It is just installed on different hosts in a scaled-up fashion to provide redundancy and high availability. Kops, Kops is a utility that installs Kubernetes, creates upgrades and deletes clusters on Amazon Web Services using the AWS command line. Kubernetes clusters are managed visually through the AWS console. So the same way that we run instance or create instances with AWS, we would create the nodes in the same way. Kops delivers an enterprise level configuration for Kubernetes. Enterprise known server requirements are enforced by Kops. So if we're going to use Kops, we have to understand that it will enforce the requirements or the enterprise requirements. Therefore, the free tier of Amazon Web Services is not an option. It will enforce the the number of processors. It will enforce levels of memory. It will enforce levels of disc. Kops natively uses AWS Route 53 and S3. Kops also requires AWS IAM roles for access. And Kops is not the method for installing Kubernetes on AWS, but it provides a very consistent approach. So, Kops is not the only method for installing Kubernetes on AWS. But again, it provides a very templated process and repeatable process. Microsoft Azure also has its own utility, AKS, and it delivers a platform for Kubernetes clusters. Like Kops, AKS provides a consistent approach for quickly creating production-ready Kubernetes clusters. AKS is a utility with a rich set of sub commands that is accessed by the Azure command line. The az aks utility allows administrators to create, delete, browse, install for Kubernetes as well as many other commands for Kubernetes support. And this is a very interesting solution. The Google Cloud platform has the Kubernetes engine, which is a tool for managing Kubernetes clusters, workloads, applications and more. As we discussed a little earlier, Kubernetes began as a Google project. Therefore, Google maintains a very rich set of tools, a rich approach for creating and building Kubernetes clusters. Google Cloud platform has a rich console and set of tools to support Kubernetes for the Enterprise. And again, Kubernetes began as a Google project, therefore, it's able to take advantage of that Google experience. Please go through the readings that are part of this lesson, and please complete the quiz.
