Book Image

Learning Windows Server Containers

Book Image

Learning Windows Server Containers

Overview of this book

Windows Server Containers are independent, isolated, manageable and portable application environments which are light weight and shippable. Decomposing your application into smaller manageable components or MicroServices helps in building scalable and distributed application environments. Windows Server Containers have a significant impact on application developers, development operations (DevOps) and infrastructure management teams. Applications can be built, shipped and deployed in a fast-paced manner on an easily manageable and updatable environment. Learning Windows Server Containers teaches you to build simple to advanced production grade container based application using Asp.Net Core, Visual Studio, Azure, Docker and PowerShell technologies. The book teaches you to build and deploy simple web applications as Windows and Hyper-V containers on Windows 10 and Windows Server 2016 on Azure. You will learn to build on top of Windows Container Base OS Images, integrate with existing images from Docker Hub, create custom images and publish to Hub. You will also learn to work with storage containers built using Volumes and SQL Server as container, create and configure custom networks, integrate with Redis Cache containers, configure continuous integration and deployment pipelines using VSTS and Git Repository. Further you can also learn to manage resources for a container, setting up monitoring and diagnostics, deploy composite container environments using Docker Compose on Windows and manage container clusters using Docker Swarm. The last chapter of the book focuses on building applications using Microsoft’s new and thinnest server platform – Nano Servers.
Table of Contents (19 chapters)
Credits
Foreword
About the Author
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface

Container resource allocation


When we create a container using docker run, the container host does not set any resource constraints on the container. If you are planning to build a multi-tenant system to host a set of microservices as containers, you would want all the services to perform well. Each service would have its own resource requirements: some services may run CPU intensive workloads, some may run memory intensive work loads, like in-memory storage, and few may bank on high bandwidth. In a multi-tenant environment or shared infrastructure, resource usage should be constrained to avoid problems like noisy neighbors. If it's not, a CPU-intensive service might eat into the host system, which impacts other services running on the same host. The following image shows a typical noisy neighbor problem:

Noisy neighbor is a common problem in cloud-computing and shared infrastructure, where a co-tenant monopolizes bandwidth, disk, IO, CPU and other resources, and can negatively impact another...