Introduction: W hile familiarizing with Cassandra, I felt the unavailability of Sub-query is polluting my application by iterating over the first query result, and the next round trip is to fetch actual data. The main reason behind is the normalized approach that I have taken while designing column families. The RDBMS style of table design is not fair for non-structured NoSQL data stores. We can keep the reference table contents in a single table as separate columns. But some use cases will not allow as keeping everything in a single column family. (Especially when we consider the performance degradation caused by the compaction when the traffic to a single column family increased since compaction are per column family based).Also if we keep the index structure in a separate table without using the inbuilt secondary index provided by Cassandra. (Separate wide row implementation of index data will allow as to perform equal, range queries against the data), two queries have to be