Spark JDBC and ODBC Drivers

Data Fabric provides JDBC and ODBC drivers so you can write SQL queries that access the Apache Spark data-processing engine. This section describes how to download the drivers, and install and configure them.

You can download the Spark JDBC driver from https://package.ezmeral.hpe.com/tools/MapR-JDBC/MapR_Spark/ and the Spark ODBC driver from https://package.ezmeral.hpe.com/tools/MapR-ODBC/MapR_Spark/.
IMPORTANT
To access the Data Fabric internet repository, you must specify the email and token of an HPE Passport account. For more information, see Using the HPE Ezmeral Token-Authenticated Internet Repository.

After downloading the driver, refer to the documentation at Spark JDBC Driver to install and configure the JDBC driver and Spark ODBC Driver for the ODBC driver. A copy of the documentation also is available in each download package. The following table describes the driver versions available for various EEP releases:

EEP version Driver version Driver link Documentation link
JDBC version EEP 6.0.0+ 2.6.3 Spark JDBC Driver for version 2.6.3 Spark JDBC Documentation for version 2.6.3
EEP 3.0.1+ 1.1.8 Spark JDBC Driver for version 1.1.8 Spark JDBC Documentation for version 1.1.8
ODBC version EEP 6.0.0+ 2.6.1 Spark ODBC Driver for version 2.6.1 Spark ODBC Documentation for version 2.6.1
EEP 2.0.2+ 1.2.5 Spark ODBC Driver for version 1.2.5 Spark ODBC Documentation for version 1.2.5
NOTE
When connecting to the Spark Thrift Server using beeline and the JDBC driver, you might encounter the following error:
"Unsupported transaction isolation level: 4"
To avoid this error, pass the isolation parameter to beeline as follows:
bin/beeline --isolation=default