Contains support for cloud infrastructures, including the Hadoop JARs and transitive dependencies needed to interact with the infrastructures. When included in the spark-assembly JAR, the hadoop artifacts are included, but not any of the 3rd party libraries needed, such as those from Amazon (for AWS) and Microsoft (Azure). These will need to be explicitly added to the classpath of any application interacting with the services. Any project which explicitly depends upon the spark-cloud artifact will get the dependencies; the exact versions of which will depend upon the hadoop version Spark was compiled against. Hadoop 2.5 and earlier: jets3t. Hadoop 2.6: hadoop-aws and aws-java-sdk JARs Hadoop 2.7+: hadoop-aws aws-java-sdk-s3 hadoop-azure azure-storage WARNING: the signatures of methods in aws-java-sdk/aws-java-sdk-s3 can change between versions: use the same version against which Hadoop was compiled.