Introduction

You can set up Knowledge Discovery on Kubernetes using Helm. Kubernetes uses the same container images as Docker or Docker Compose, but allows complex applications to be deployed across multiple machines and automatically scale in response to user demand.

Helm is a package manager for Kubernetes that allows an entire solution, that can be deployed in Kubernetes, to be defined in a chart. One or more charts can be made available in a helm repository. The idol-containers-toolkit includes a helm repository containing helm charts for deploying some example Knowledge Discovery solutions.

One of the charts provided in the idol-containers-toolkit is the distributed-idol chart. This deploys a DAH, DIH, and a set of Content components that can autoscale with demand. The chart can deploy in mirror-mode (scaling to fulfill query demand) or in non-mirror mode (scaling to fulfill index demand) depending on the setupMirrored Helm value. The default is non-mirror mode. For more information about DAH, DIH, and Content, see Knowledge Discovery Core Components.

NOTE: This guide assumes that you have installed and set up the Kubernetes platform in your environment. It does not provide general information about how to use Kubernetes. For more information about Kubernetes, refer to the Kubernetes documentation.

If you want to set up a Kubernetes cluster on a single machine for learning or demonstration purposes, you can use k3d.