Orchestration Basics

While containers solve the problem of packaging and running applications consistently, managing a large number of containers across a fleet of machines presents a new challenge. This is where container orchestration comes in. Orchestration platforms automate the deployment, management, scaling, and networking of containers. The undisputed leader in this space is Kubernetes, an open-source project originally developed by Google. Kubernetes provides a powerful framework for running distributed systems resiliently. It takes care of scaling your applications up or down based on demand, ensures they restart if they fail (self-healing), manages how they talk to each other, and handles rolling updates with no downtime. A user simply declares the desired state of their application—for example, 'I want to run 3 replicas of my web server container and expose it to the internet'—and Kubernetes works to make the current state match the desired state. It abstracts away the underlying host machines, allowing developers to treat an entire cluster of servers as a single, massive computational resource. Other orchestration tools include Docker Swarm and Apache Mesos, but Kubernetes has become the de facto standard for managing containerized applications at scale in the cloud.