Connect to a CDH4.5 Cluster

Question

I have a CDH4.5 cluster, and I want to upload files into it from another server (eg database server).

With vanilla Hadoop and Hive, I can change the configuration files, pointing the namenode and metastore to remote services, and simply run:

dba@db-001$ hadoop fs -copyFromLocal /path/to/export.tsv
dba@db-001$ hive -e "load data local inpath '/path/to/export.tsv' into table test.my_table"

But how about CDH? What components should I install on other servers?

Answer 1

1. Download hadoop and hive tarballs corresponding to the CDH version: http://www.cloudera.com/content/cloudera-content/cloudera-docs/CDH4/latest/CDH-Version-and-Packaging-Information/cdhvd_topic_3.html#topic_3_1_unique_7

2. Download configuration files from cloudera manager, unzip them, and point HADOOP_CONF_DIR/HIVE_CONF_DIR to these folders.