With the Cassandra database system, is it okay to use a boolean valued column as a clustering column (as part of the primary key)? The column iwll of course not be a partiti