On the exam you should always use Ambari when possible, especially for tasks like enabling NameNode HA. ... View more

@John Cod - Try adding the following environment variables to spark-env.sh (found in the /conf folder of your Spark install) - using the appropriate IP address of course if Spark is running on a machine other than localhost: export SPARK_MASTER_IP=127.0.0.1 export SPARK_LOCAL_IP=127.0.0.1 ... View more

@Anshu Kumar - this is not a concern that you need to have on the exam. In order for a task on the exam to be marked as correct, you must have already executed your code and the output must exist in HDFS. If the desired output folder specified by the task instructions is not in HDFS, then that task will be immediately marked as wrong. Do not worry about how the evaluator grades the exam. Focus on generating the desired output by carefully following the task instructions, making sure that you run your code, and making sure your code generated the desired output. ... View more