Running H2O cluster on a Kubernetes cluster - Open Source Leader in AI and ML

H2O is an open-source, in-memory platform for distributed, scalable machine learning. With the major release 3.30.0.1, released in Q1 2020, H2O obtained first class Kubernetes support. This article explains how to create H2O deployment on Kubernetes. It also covers the selected H2O internal mechanisms for the reader's better understanding of H2O's behavior on Kubernetes cluster. In order to understand the behavior and limitations of H2O distributed cluster, it is mandatory to understand the basics of H2O design.