Utilizing the column storage of Apache Cassandra

Apache Cassandra is a column-family NoSQL database that can also hold large amounts of data. HBase can share big data across its regions using auto-sharding, which makes HBase horizontally scalable. Likewise, Cassandra supports adding more nodes horizontally to improve server throughput, a characteristic of horizontal scaling. But there are also some differences between the two storages in terms of their architectures, table read and write performances, data modeling approaches, and query languages.

Let’s start by designing our course, degree_level, student, and student_perf Cassandra tables.

Designing Cassandra tables

Unlike HBase, Cassandra stores its data per row, grouping all column fields, thus why the data model approach is a column family. Its database transactions are atomic, isolated, and durable, but with eventual or tunable consistency, so it doesn’t offer an Atomicity, Consistency, Isolation, Durability...