To enable an Apache NiFi data flow to communicate with Hive, you must obtain the Hive
connection details by downloading several client configuration files. The NiFi processors
require these files to parse the configuration values and use those values to communicate
with Hive.
-
From the Services pane in the Data Engineering cluster
that contains the Hive instance to which you want to ingest data, click
CM-UI.
-
Click Hive Metastore, located on the Cloudera Manager
Clusters pane.
-
From the Hive Metastore view, click to download the client configuration files as a zip file.
-
Expand the zip file and upload the following files to each of the NiFi nodes in
your Flow Management cluster:
- hdfs-site.xml
- hive-site.xml
- core-site.xml
For example:
scp hdfs-site.xml cdp_username@publiclp:/tmp
You may also copy these files to a more permanent location later.
-
Adjust the file permissions to ensure that the NiFi processors can read them.
For
example:
chmod 755 /tmp/*-site.xml