Are column families within the same key space in any way?

Question

What is the difference between having Foo space keys and columns A and B in it versus having two clusters FooA and FooB with one column family in each?

The API looks as if the two were pretty much equivalent.

As a bonus question, how do supercolumns fit into this picture?

Answer 1

Keyspace : The namespace for ColumnFamilies, usually one for each application. The key space is the first dimension of the Cassandra hash and is a container for column families. Classes have roughly the same granularity as a schema or database (i.e., a logical set of tables) in the RDBMS world. This is the configuration and management point for column families, as well as the structure in which package insertions are applied.

ColumnFamilies contains several columns, each of which has a name, value, and timestamp and referenced by row keys.

Supercolumns can be thought of as columns that themselves have sub-columns (columnfamily inside columnfamily).

A finer explanation of Cassandra's data model is found here.