Uploading additional JARs to CDW
You add additional Java Archive (JAR) files to the Cloudera Data Warehouse (CDW) Hive classpath that might be required to support dependency JARs, third-party Serde, or any Hive extensions.
- The JARs are added to the end of the Hive classpath and do not override the Hive JARs.
- Cloudera recommends that you do not use this procedure to add User-Defined Function (UDF) JARs. If you do, then you must restart HiveServer2 or reload the UDF. For more information about reloading functions, see the Hive Data Definition Language (DDL) manual.
Build the archive file.
The archive file can be either a .jar file or a tar.gz file. For a tar.gz archive file, only JARs present in the top level are considered.
For example, if the tar.gz file contains these files — test1.jar, test2.jar, and deps\test3.jar, only test1.jar and test2.jar are considered; deps\test3.jar is excluded.
- Upload the archive file to the Hive Virtual Warehouse on CDW object storage, such as HDFS.
- Log in to the CDW service and from the Overview page, locate the Hive Virtual Warehouse that uses the bucket or container where you placed the archive file, and click and select Edit.
- In the Virtual Warehouse Details page, click .
- From the Configuration files drop-down list, select env.
CDW_HIVE_AUX_JARS_PATHand add the archive file to the environment variable.
If you add a directory, the .jar or tar.gz files within the directory are copied and extracted. For a tar.gz file, only the JARs present in the top level are copied.Consider the following JAR path - /common-jars/common-jars.tar.gz:/common-jars/single-jar.jar:/serde-specific-jar/serde.jar. In this example,
common-jars.tar.gzis extracted and
serde.jarfiles are copied.
Repeat the previous step and add the archive file or directory for
Query coordinator and Query
CDW_HIVE_AUX_JARS_PATHenvironment variable is not present, click and add the following custom configuration: