A brief description of data modeling in Cassandra.
At one level, Cassandra tables, rows, and columns can be thought of much the same way as those in a relational database. In both SQL and CQL you define tables, which have defined columns and associated data types, and you can create indexes to allow efficient querying by column values.
However, an important difference is that since Cassandra is designed from the ground up as a distributed system, it emphasizes denormalization instead of normalization and joins, and provides tools like collections to support this.
Attention: Be sure this document version matches your product version!