Play with Hadoop
Settings up a Single Node Hadoop Cluster 1.0.3
See a wrapper cookbook for a smoother start.
- Ubuntu
Tested on:
- Linux precise64 3.2.0-23-generic #36-Ubuntu SMP x86_64 GNU/Linux
- Hadoop requires Oracle 1.6 JDK/JRE (see http:https://wiki.apache.org/hadoop/HadoopJavaVersions). See the attirbute section below for
node['single_node_hadoop_claster']['java']['java_home']
attribute
node['single_node_hadoop_claster']['java']['java_home']
- path to JAVA_HOME (for example: "/usr/lib/jvm/jdk1.6.0_45").node["single_node_hadoop_claster"]["user"]["name"]
- hdfs/hadoop user. Note, that the cookbook doesn't distinguish between these two roles (hdfs user and map/reduce user).node["single_node_hadoop_claster"]["user"]["group"]
- hdfs/hadoop user group.
See the attributes/default.rb
for more info.
- See https://github.com/MitinPavel/single_node_hadoop_claster_wrapper It is a wrapper cookbook, which configures Apt and Java before starting this cookbook.
- Author:: Pavel Mitin ([email protected])
Copyright 2013 Pavel Mitin
Licensed under the MIT License (MIT).