Hive supports column based storage using RC files. https://issues.apache.org/jira/browse/HIVE-352
The question is: if I store hive tables in this format and once in a while would like run some custom map-reduce jobs against the data in this table, am I restricted in any sense? Thanks, Abhi
