The File System API can read data from HDFS, S3, and Local FileSystem, among other storage devices. Spark uses the FS API to get information from different storage engines.
