Name: puppet-hadoop
Owner: The Hyve
Description: Hadoop Module for Basic Installation of Core/Common Package
Forked from: swizzley/puppet-hadoop
Created: 2016-10-24 12:30:21.0
Updated: 2016-10-24 12:30:23.0
Pushed: 2016-10-24 12:40:28.0
Homepage: null
Size: 20
Language: Puppet
GitHub Committers
User | Most Recent Commit | # Commits |
---|
Other Committers
User | Most Recent Commit | # Commits |
---|
Table of Contents
Hadoop Module for setup of Core/Common Hadoop Package on Linux systems.
This could take a while to run, the current version is 200MB, others are smaller and you can specify which to use in params as well as the url for the mirror that works best. The latest version of artifact has no exec timeout, so that the file will eventually download.
ude ::hadoop
New fact $::java_home added for finding and setting $JAVA_HOME
$version
: [default: 2.7.1] which is current stable at the time of this module release
$package
: Specify Core or Common Package [default: common]
target
: Download directory [default: /tmp]
install_dir
: Installation path [default: /opt/hadoop]
java_dist
: Java installation for use when installing java via puppetlabs/java [default: undef] which is effectively jdk
java_pkg
: Specify a java package to pass to puppetlabs/java [default: undef]
java_ver
: Specify a java version to pass to puppetlabs/java [default: 'present']
owner
: Owner of hadoop directory and files [default: hadoop]
group
: Group of hadoop directory and files [default: hadoop]
Linux:
Tested On: CentOS 6.6 (used in production)
If installing java using this module which uses the puppetlabs/java module, puppet will need to run twice before $JAVA_HOME can be set to the valid value of $::java_home, because facter gets read before java is installed. If java is already installed, it will work the first run.
I'd like to add zookeeper, hbase, hive, pig and other parameters and manifests, but there are other modules for that already, and the cloudera modules is already puppet approved, so we'll just see where this goes.