Network infrastructure management and operations are often very manually intensive processes, prone to error and slow to react to business needs. DevOps is a set of practices and tools in software development, focusing on efficiency to integrate and automate development, testing, and deployment to production in order to improve the software systems development life cycle. In this book, Carter and King explain how to apply the approach of DevOps to physical network infrastructure operations in order to automate and increase agility, security, and reliability. The goals and layout of the book are described in the preface, along with an explanation of the use of a fictional company throughout the book to help illustrate the concepts.

In chapter 1, Carter and King discuss the problems with legacy mechanisms for operating network infrastructure and how applying the principles of DevOps could address them. Chapter 2 provides a more detailed discussion on the issues involved in managing network infrastructure “box-by-box”; supported with diagrams and code examples, the authors explore the alternative of applying model-driven DevOps to simplify network infrastructure management. This process uses standard data models and a platform to organize and transform these high-level models into complex device configurations. Chapter 3 introduces and discusses the concept of consumable infrastructure, which is infrastructure that can be accessed via application programming interfaces (APIs), using data models and a platform to transform the high-level data model into specific API calls to a variety of different physical infrastructures.

Chapter 4 introduces the concept of infrastructure as code (IaC), describing the pipelining of configuration data from a “source of truth” data store through device-specific transforms on a management platform and on to the devices in question using APIs. Common IaC tools and the importance of structured organization of configuration data are discussed with detailed code examples. In essence, IaC allows for the automation of network infrastructure management at scale--although it can also magnify the impact of any mistakes. Importantly, the platform manages the order in which changes are applied. Chapter 5 continues this discussion to the continuous integration/continuous deployment (CI/CD) concept, also from DevOps, that extends the flow through a pipeline of coding, testing, and deployment that will be familiar to application developers. Concepts such as source code management, version control, code branching, issue tracking, and testing on infrastructure simulators are covered in detail.

Chapter 6 then provides details of how model-driven DevOps tools for network infrastructure might be implemented, including full code for a reference implementation, provided in a GitHub repository. Chapter 7 looks at the weakest link in all information technology (IT) systems: the human factors. Discussion covers approaches to driving cultural change and the leadership needed to adapt the culture within an organization, that is, leaders who can explain the need for, and benefits of, the model-driven DevOps approach proposed here.

Throughout the book, the exploits of Bob, a fictional character in the generic ACME Corporation, are effectively used to illustrate the authors’ points. Indeed, many of the problems faced by Bob bring back frightening memories from my own career. Carter and King provide the reasoning behind the need for model-driven DevOps; describe how to use it to automate and increase agility, security, and reliability of network operations; and provide significant detail on how to implement such a solution. There is a good table of contents and a thorough index, and each chapter is illustrated with detailed code examples. This book will be invaluable for IT infrastructure teams wanting to adapt, improve, and automate their management practices for network infrastructure.

More reviews about this item: Amazon