Wide column databases are NoSQL databases that store data in tables, rows, and columns. The name and column formats can vary from one row to another. Wide column databases are generally interpreted as 2D key-value or key key-value stores. Wide column databases are also known as column family databases. These databases are used for transactional workflows. Wide column databases can be used in cases with large datasets distributed across multiple database nodes. Given big data volumes in IoT, data scientists usually use wide column databases to store IoT and user profile data. It is also used to manage big data warehouses because of its ability to scale.
Wide column databases help horizontal scalability of data and provide data consistency. Data is stored in column cells which are grouped into column families. Wide column databases can filter data by value. Wide column databases differ from columnar databases with respect to storage of data. Columnar databases store each column separately on a disk, whereas wide column databases support column families stored together. Wide column databases also differ from relational databases. Relational databases have multiple tables whereas wide column databases have multiple column families.
To qualify for inclusion in the Wide Column Databases category, a product must:
Contain 2D key-value or key key-value
Have clustering for distributed database
Provide time to live (TTL) management
Support a variety of data types like String, Boolean, Integer, and Double
Provide a three dimensional structure (row, column, and time)
