I\'m running a Spark cluster in EMR. All of the execution works fine but, in a specific stage, it returns to process previows stages showing retry in the console:
