Installation

Project Koku is divided into logical components. To achieve a fully functional application, it requires the deployment of multiple micro-applications.

The primary applications for Project Koku are:

  • Koku API (Reporting and Query API service)
  • Koku UI (Front-end Web UI)
  • Masu (Data ingestion service)

For development and testing, Project Koku also includes:

  • Nise (Test data generator)

This guide will focus on deploying Project Koku into an existing OpenShift cluster.

Deploying the Koku API

The Koku application contains two components - a web service and database.

OpenShift

A basic deployment configuration is contained within the application’s openshift template file. This template should be acceptable for most use cases. It provides parameterized values for most configuration options.

To deploy the Koku API application using the provided templates, you can use the provided Makefile: (the openshift CLI is required to be installed to run the following command)

make oc-create-all

To deploy individual components, there are also make commands provided for your convenience:

Deploy the API web application: make oc-create-koku Deploy the PostgreSQL database: make oc-create-db

Docker Compose

The Koku API can also be deployed with Docker Compose with the following steps:

  • Create a Docker bridge network named koku-network: docker network create koku-network
  • Set AWS credential environment variables: AWS_SECRET_ACCESS_KEY and AWS_ACCESS_KEY_ID
  • Start koku server and database: make docker-up

This command will run database migraitons and start the API server. Once complete the API server will be running on port 8000 on your localhost.

Deploying the Koku UI

The Koku-UI application is the web-based frontend for Project Koku. It is built using Patternfly and ReactJS.

Instructions for deploying the Koku UI can be found here: https://github.com/project-koku/koku-ui#getting-started

Deploying Masu

The Masu application contains several components - a web service, message bus, and workers. Masu also uses the Koku database. Configuration and management of the database are controlled from the Koku API application.

Instructions for deploying Masu can be found here: https://github.com/project-koku/masu#getting-started

Masu can alternatively be deployed with Docker Compose: Follow the steps above to start the Koku API with Docker Compose

Start masu services: masu_branch> make docker-up To trigger Masu to process data, send a GET request to http://<baseUrl>:5000/api/cost-management/v1/download/ You may want to see Masu logs using docker logs koku_worker -f