Book Image

The DevOps 2.1 Toolkit: Docker Swarm

By : Viktor Farcic
Book Image

The DevOps 2.1 Toolkit: Docker Swarm

By: Viktor Farcic

Overview of this book

Viktor Farcic's latest book, The DevOps 2.1 Toolkit: Docker Swarm, takes you deeper into one of the major subjects of his international best seller, The DevOps 2.0 Toolkit, and shows you how to successfully integrate Docker Swarm into your DevOps toolset. Viktor shares with you his expert knowledge in all aspects of building, testing, deploying, and monitoring services inside Docker Swarm clusters. You'll go through all the tools required for running a cluster. You'll travel through the whole process with clusters running locally on a laptop. Once you're confident with that outcome, Viktor shows you how to translate your experience to different hosting providers like AWS, Azure, and DigitalOcean. Viktor has updated his DevOps 2.0 framework in this book to use the latest and greatest features and techniques introduced in Docker. We'll go through many practices and even more tools. While there will be a lot of theory, this is a hands-on book. You won't be able to complete it by reading it on the metro on your way to work. You'll have to read this book while in front of the computer and get your hands dirty.
Table of Contents (22 chapters)
Title Page
Credits
About the Author
www.PacktPub.com
Customer Feedback
Preface
11
Embracing Destruction: Pets versus Cattle

Clustering


A server cluster consists of a set of connected servers that work together and can be seen as a single system. They are usually connected to a fast Local Area Network (LAN). The significant difference between a cluster and a group of servers is that the cluster acts as a single system trying to provide high availability, load balancing, and parallel processing.

If we deploy applications, or services, to individually managed servers and treat them as separate units, the utilization of resources is sub-optimum. We cannot know in advance which group of services should be deployed to a server and utilize resources to their maximum. Moreover, resource usage tends to fluctuate. While, in the morning, some service might require a lot of memory, during the afternoon that usage might be lower. Having predefined servers prevents us from having elasticity that would balance that usage in the best possible way. Even if such a high level of dynamism is not required, predefined servers tend...