How amazon-archives/emr-sample-apps works
Project type and purpose
• This is a “sample code” repository illustrating how to run different data-processing frameworks and tools on Amazon EMR.
• It contains multiple independent modules (cloudburst, cloudfront, freebase, similarity, etc.), each demonstrating a particular EMR use case (Java MapReduce, Cascading, Python/MRJob, Pig, Hive-Ads, Spark/Shark, Node.js MR, encryption plugin for EMRFS, etc.).
File structure
• Top-level directories each represent an independent sample application or plugin.
• cloudburst: Java-based MapReduce alignment tool
• cloudfront: Java/Cascading log processor
• emrfs-plugins: EMRFS encryption provider
• freebase, similarity: Python/Ruby MRJob or Hadoop streaming jobs
• hive-ads, pig-apache: Hive and Pig scripts
• spark: Spark/Shark query samples
• node, wordcount: Hadoop streaming with Node.js / Python
• powershell: Windows automation scripts for Hive jobs
• No single interdependency across modules—they share the common target environment (Amazon EMR + S3 + SimpleDB).
README for additional insights
• The README states: “code samples … illustrate how to use popular applications on Amazon EMR.”
Detail without overthinking
• Logical EMR component and its responsibility.
• That all “sample modules” are just different job steps in the same overarching architecture.