We are rolling out Hadoop nodes. I'd prefer to keep and manage the configuration for each system outside of the HADOOP_HOME location. There is HADOOP_CONF_DIR which can be set - but it's not really clear to me, after reading the scripts and some docs, where the best place to set this is, so that everything that needs to reference HADOOP_CONF_DIR will do so correctly.
For example, there the hadoop-env.sh script, the init.d scripts, etc.
Can someone point me in the right direction. Thanks.
[sup][/sup]
For example, there the hadoop-env.sh script, the init.d scripts, etc.
Can someone point me in the right direction. Thanks.
[sup][/sup]