BMQ Data Science Bundle

Description

This bundle contains blueprints and recipes necessary to setup a short lasting cluster training ML pipeline for the Beast Mode Quotient application detailed here: https://community.hortonworks.com/content/kbentry/226140/beast-mode-quotient-using-hybrid-cloud-architectur.html

How-To

• Preparation

  • Setup an instance of Cloudbreak with the right credentials
  • Change the tp-bmq-data-science.json to reflect these credentials
  • Add the Blueprint bd-bmq-data-science.json to your CB instance
  • Add the pre-ambari recipe pras-bmq-data-science.sh to your instance
  • Install CB cli
  • Change launch-bmq-ephemeral-cluster.sh and terminate-bmq-ephemeral-cluster.sh to reflect your CB location, username, password as well as the location of the cli

• Launch an ephemeral cluster

  • Run launch-bmq-ephemeral-cluster.sh

• Terminate all launched clusters

  • Run terminate-bmq-ephemeral-cluster.sh

Versions

• Cloudbreak 2.8.0
• HDP 3.1

Authors

• Paul Vidal - Initial work - LinkedIn