Configure Connect for Big Data Sort - Connect_ETL - 9.13

To invoke the Connect for Big Data sort accelerator for a MapReduce job, the following configuration settings are required:

• The dmxhadoop jar file that contains the Connect for Big Data Sort classes must be in the Java CLASSPATH on all nodes in the cluster in order for the Sort accelerator to be invoked. The file can be specified as:

<connect_install>/lib/dmxhadoop_<type>.jar

where <type> is mrv1 or mrv2 (mrv1_mapr or mrv2_mapr for MapR) as per the MapReduce version running in your cluster.

• The required Connect for Big Data sort accelerator properties must be set as indicated in the following section.