##Druid Puppet Module

This is a basic puppet module for managing various Druid nodes. The module will download and install the defined Druid version libraries, bootstrap the defined Druid nodes, and maintain the Druid server processes using Supervisord.

WTF is Druid? Druid is an open-source analytics data store designed for real-time exploratory queries on large-scale event data. Druid provides cost-effective, highly-available real-time data ingestion and arbitrary data exploration. Built by a few really awesome guys over at Metamarkets.

Disclaimer: This is a very rough module and has only been tested so far with Ubuntu 12.04 LTS 64bit, but should work with most Debian based distributions. I would recommend only using this module in development unless you really know what your doing and have read through all of the Druid.io Docs.

---

###Requirements:

• Puppet stdlib
• Puppet supervisor
• Puppet concat
• JRE (will install openjdk-7-jre-headless by default)
• AmazonS3 account for historical nodes
• Available ZooKeeper instance

---

###Todo:

• DRY things up
• Better support for hiera
• Dynamic nodes
• Better indexing service support
• Better supervisord handling
• Hadoop indexing configurations
• Better download handling
• More validations
• Pre-configuration to run out of the box
• Optional S3 (may not need historical)
• Other platform support
• Declarative RT node spec configuration
• Vagrantfile
• Example manifests for default Druid examples

NOTE: Using the default Realtime Node configuration, you will be required to have a RabbitMQ instance available. See the extensions property for Class druid::node::realtime.

---

###Bootstrap

  class { 'druid':
    mysql_host        => 'localhost',
    mysql_port        => 3306,
    mysql_db          => 'druid',
    mysql_user        => 'root',
    mysql_pass        => 'root',
    storage_bucket    => 'some_awesome_bucket',
    storage_base_key  => 'some_bucket_base_key',
    s3_access_key     => 'SOME_KEY',
    s3_secret_key     => 'SOME_SECRET',
  }

---

###All nodes on single server

  # Druid Coordinator (cluster management)
  class { 'druid::node::coordinator':
    port  => 8090,
  }
  
  # Druid Historical (deep storage)
  class { 'druid::node::historical':
    port => 8091,
  }

  # Druid Realtime (event processing)
  class { 'druid::node::realtime':
    port        => 8092,
    spec_hash   => [{ ...my realtime spec... }],
  }

  # Druid Broker (query broker)
  class { 'druid::node::broker':
    port => 8093,
  }

  # Druid Overlord (indexing service)
  class { 'druid::node::overlord':
    port => 8094,
  }

---

###License

GNU General Public License v3 (GPL-3)