The document discusses different ways to access a Hadoop cluster running on MapR technologies. It describes the MapR Control System (MCS) for administrative access via a web interface, the MapR Client for submitting jobs from a local machine, accessing the cluster filesystem via NFS for drag-and-drop of files, and using the Hive ODBC connector to query data with BI tools. Exercises are provided for hands-on experience with each of these access methods.