Hybrid NoSQL Databases

By Adam Fowler

Given the range of data types being managed by NoSQL databases, you’re forgiven if you think you need three different databases to manage all your data. However, although each NoSQL database has its core audience, several can be used to manage two or more data structures. Some even provide search on top of this core all‐data platform.

Hybrid databases can easily handle document and key‐value storage needs, while also allowing fast aggregate operations similar to how column stores work. Typically, this goal is achieved by using search engine term indexes, rather than tabular field indexes within a table column in the database schema design itself.

The functionality provided, though, is often the same as in column stores. So, these products have three or four of the preceding types covered: key‐value, document, and column stores, as well as search engines.

Many databases are moving in this direction.