Book Image

Cloud Native with Kubernetes

By : Alexander Raul
Book Image

Cloud Native with Kubernetes

By: Alexander Raul

Overview of this book

Kubernetes is a modern cloud native container orchestration tool and one of the most popular open source projects worldwide. In addition to the technology being powerful and highly flexible, Kubernetes engineers are in high demand across the industry. This book is a comprehensive guide to deploying, securing, and operating modern cloud native applications on Kubernetes. From the fundamentals to Kubernetes best practices, the book covers essential aspects of configuring applications. You’ll even explore real-world techniques for running clusters in production, tips for setting up observability for cluster resources, and valuable troubleshooting techniques. Finally, you’ll learn how to extend and customize Kubernetes, as well as gaining tips for deploying service meshes, serverless tooling, and more on your cluster. By the end of this Kubernetes book, you’ll be equipped with the tools you need to confidently run and extend modern applications on Kubernetes.
Table of Contents (22 chapters)
1
Section 1: Setting Up Kubernetes
5
Section 2: Configuring and Deploying Applications on Kubernetes
11
Section 3: Running Kubernetes in Production
16
Section 4: Extending Kubernetes

Chapter 11 – Template Code Generation and CI/CD on Kubernetes

  1. Helm Charts use templates and variables, while Kustomize uses a patch-based strategy. Kustomize is built into recent versions of kubectl, while Helm uses a separate CLI tool.
  2. The configuration should emphasize security, since deploy credentials could be used to deploy attacker workloads to your cluster. Using either secure environment variables or access management controls on your cloud provider are two good strategies. The credentials should absolutely not be placed in any Git repository.
  3. In-cluster setups can be preferable since Kubernetes credentials are not required to be provided by an external system. Out-of-cluster setups are usually simpler, and more synchronous than in-cluster setups, where a control loop determines when changes are made to the resource configuration.