Chapter 6: Kubernetes Container Orchestration Infrastructure Management
While it is fairly straightforward to build a local data science environment with open source technologies for individual uses in simple machine learning (ML) tasks, it is quite challenging to configure and maintain a data science environment for many users for different ML tasks and track ML experiments. Building an end-to-end ML platform is a complex process, and there are many different architecture patterns and open source technologies available to help. In this chapter, we will cover Kubernetes, an open source container orchestration platform that can serve as the foundational infrastructure for building open source ML platforms. We will discuss the core concept of Kubernetes, its networking architecture and components, and its security and access control. You will also get hands-on with Kubernetes to build a Kubernetes cluster and use it to deploy containerized applications.
Specifically, we will cover...