The three products provide interfaces at different levels to large data
|Hadoop File System||Used for storing Organized Raw Data as Blobs|
|Hive||Used to run analytics on entire data-sets, and produce fresh data-sets|
|HBase||A basic tabular data-store capable of storing large amounts of data|
Hadoop Distributed File System
Hadoop Distributed File System (HDFS) provides the lower level service (Blob Abstraction) on top of which HBase and Hive provide a row-based abstraction.
Hive works more on stable data-sets. It has Write capability, but this is to create new data-sets by processing old data-sets, or to create new versions of the same data-set.
Hive is well suited for OLAP Analytics, and for recording Transaction History.