Create a local working CDH5 Hadoop with ansible scripts
These ansible scripts accompany this GoDataDriven blog about installing a local CDH5 hadoop in local and pseudo distributed mode.
If not done yet, install ansible by e.g.
$ brew install ansible
Start installing by issuing:
$ start-playbook.sh
This script will ask for the directory to install into. After that the ansible playbook is started which will:
- download the tarballs (hadoop, hive, pig, mahout and spark currently)
- unpack them into the install directory
- creates a .bash_profile_hadoop to setup the aliasses
- configures the hadoop and hive config files
- fixes a few lines in hadoop-config.sh to make sure JAVA_HOME is set correctly
TODO: