Data on these systems are accessed via Java API or the Hadoop command line client. Most of the operations you find in Hadoop are similar to their Unix counterparts.