Various samples with Spark
- build your application into a .jar file
- make sure all dependencies are assembled into jar file
- upload it to Hadoop master node
- use following command to submit it
spark-submit --class com.spark.examples.worstMovies.worstMoviesSPOY.WorstMoviesSPOY --master yarn --deploy-mode cluster /media/shared_from_local/sparkexamples.jar
(replace class name and jar with yours) - logs can be traced via following command
yarn logs -applicationId application_1538675682869_0113
(replace with your application id retrieved from spark-submit output) - Job status can be tracked through Hue or Spark UI