In the Repository tree view, the Hadoop cluster node in the Metadata folder groups under it the metadata of the connections to the Hadoop elements such as HDFS, Hive or HBase. It allows you to centralize the connection properties you set for a given Hadoop distribution and then to reuse those properties to create separate connections to each Hadoop element.
Click Metadata in the Repository tree view to expand the relevant folder. Each of the connection nodes will gather the various connections and schemas you have set up. Among these connection nodes is theHadoop cluster node.
The following sections explain in detail how to use the Hadoop cluster node to set up:
an HBase connection,
an HCatalog connection,
an HDFS file schema,
a Hive connection, and
an Oozie connection.
If you need to create a connection to Cloudera's analytic database, Impala, you need to use the DB connection node under the Metadata node of the Repository. Its configuration is similar to that of a Hive connection but less complicated than the latter.
For further information about this DB connection node, see Centralizing database metadata.