Book Image

The Kubernetes Bible

By : Nassim Kebbani, Piotr Tylenda, Russ McKendrick
4 (3)
Book Image

The Kubernetes Bible

4 (3)
By: Nassim Kebbani, Piotr Tylenda, Russ McKendrick

Overview of this book

With its broad adoption across various industries, Kubernetes is helping engineers with the orchestration and automation of container deployments on a large scale, making it the leading container orchestration system and the most popular choice for running containerized applications. This Kubernetes book starts with an introduction to Kubernetes and containerization, covering the setup of your local development environment and the roles of the most important Kubernetes components. Along with covering the core concepts necessary to make the most of your infrastructure, this book will also help you get acquainted with the fundamentals of Kubernetes. As you advance, you'll learn how to manage Kubernetes clusters on cloud platforms, such as Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP), and develop and deploy real-world applications in Kubernetes using practical examples. Additionally, you'll get to grips with managing microservices along with best practices. By the end of this book, you'll be equipped with battle-tested knowledge of advanced Kubernetes topics, such as scheduling of Pods and managing incoming traffic to the cluster, and be ready to work with Kubernetes on cloud platforms.
Table of Contents (28 chapters)
1
Section 1: Introducing Kubernetes
5
Section 2: Diving into Kubernetes Core Concepts
12
Section 3: Using Managed Pods with Controllers
17
Section 4: Deploying Kubernetes on the Cloud
21
Section 5: Advanced Kubernetes

The kube-scheduler component

The kube-scheduler component is a control plane component. It should run on the master node.

This is a component that is responsible for electing a worker node out of those available to run a newly created pod.

The role of the kube-scheduler component

Similar to the Kubelet, kube-scheduler queries the kube-apiserver at regular intervals in order to list the pods that have not been scheduled. At creation, pods are not scheduled, which means that no worker node has been elected to run them. A pod that is not scheduled will be registered in Etcd but without any worker node assigned to it. Therefore, no running Kubelet will ever be aware that this pod needs to get launched, and ultimately, no container described in the pod specification will ever run.

Internally, the pod object, as it is stored in Etcd, has a property called nodeName. As the name suggests, this property should contain the name of the worker node that will host the pod. When this...