deb-sahara/etc/edp-examples/edp-java/oozie_command_line
Demid Dementev 7d317bc0bc Fixed typo in the Oozie CL documentation
Change-Id: I084c8d304d3568f0b918cc37f3ae65e1d5769ece
2015-06-02 16:23:29 +01:00
..
wordcount Create etc/edp-examples directory 2014-08-13 14:11:02 +00:00
README.rst Fixed typo in the Oozie CL documentation 2015-06-02 16:23:29 +01:00

Running WordCount example from the Oozie command line

  1. Copy the edp-java.jar file from sahara/edp-examples/edp-java to ./wordcount/lib/edp-java.jar.
  2. Modify the job.properties file to specify the correct jobTracker and nameNode addresses for your cluster.
  3. Modify the workflow.xml file to contain the correct input and output paths. These paths may be sahara swift urls or HDFS paths.
    • If swift urls are used, set the fs.swift.service.sahara.username and fs.swift.service.sahara.password properties in the <configuration> section.
  4. Upload your wordcount directory to the oozie.wf.application.path HDFS directory (the oozie.wf.application.path directory is specified in the job.properties file):

$ hadoop fs -put wordcount oozie.wf.application.path

  1. Launch the job, specifying the correct oozie server and port:

$ oozie job -oozie http://oozie_server:port/oozie -config wordcount/job.properties -run

  1. Don't forget to create your swift input path! A sahara swift url looks like swift://container.sahara/object.