Overview/Description
An architecture is considered resilient if it is continuously operational and can sustain failures. Kubernetes high availability is all about setting up Kubernetes, along with its supporting components, in a way that leaves no single point of failure, and has the capability to detect hardware or software faults and remediate them. In this course, you’ll learn the Kubeadm commands and flags that can be used to manage, bootstrap, and join Kubernetes clusters. You’ll explore the highly-available Kubernetes architecture, the benefits of multi-master HA architecture, and the advantages and disadvantages of approaches for setting up HA Kubernetes clusters. Next, you’ll investigate the stacked and external etcd topologies, the role of etcd in Kubernetes, and the concepts of leaders and elections. You’ll learn about the essential control plane components and how to back up etcd clusters and use them to recover Kubernetes clusters. You'll examine how to create a load balancer for kube-apiserver and add control plane nodes to it, initialize a stacked control plane, and join multiple stacked control plane nodes. You’ll discover how to set up HA clusters with external etcd nodes, add additional control planes to the clusters, install workers after bootstrapping a control plane, and finally take snapshots using etcdctl commands and use the snapshots to restore clusters. This course is part of a series that aligns with the objectives for the Certified Kubernetes Administrator exam and can be used to prepare for this exam.
list kubeadm commands and flags that can be used to manage Kubernetes clusters, bootstrap clusters, join existing clusters, and manage tokens on running clusters
recall the architecture of highly available Kubernetes and the benefits of multi-master high availability architecture
recognize the advantages and disadvantages of the different approaches for setting up highly available Kubernetes clusters
recall the stacked and external etcd topologies that are used to set up highly available Kubernetes clusters
list the essential components of the control plane and their associated features
create a load balancer for kube-apiserver and add control plane nodes to the load balancer
initialize a stacked control plane by applying a CNI plugin and watch the pods of the control plane components
join multiple control plane nodes in parallel to the first control plane node created using the stacked control plane approach
set up highly available clusters with external etcd nodes and add additional control planes to the clusters
install workers after bootstrapping a control plane and demonstrate the steps involved in manual certificate distribution
recognize the role of etcd in Kubernetes, the concept of leaders and elections, and the terms used by etcd-based clusters for leader selection
create a three node etcd cluster using the static bootstrapping mechanism
recall approaches of backing up etcd clusters that can be used to recover Kubernetes clusters under critical disaster
work with etcdctl commands to take snapshots from live members and use them to restore clusters