Flume is a distributed, reliable, and available service for efficiently collecting, aggregating, and moving large amounts of log data. It has a simple and flexible architecture based on streaming data flows. It is robust and fault tolerant with tunable reliability mechanisms and many failover and recovery mechanisms.

module.exports = ->
  'configure': [
  'install': [
    'ryba/hadoop/core' # Users and groups created by "zookeeper" and "hadoop-hdfs" dependencies