Combined types combine SQL and NoSQL approaches and include NewSQL and multidimensional solutions.
NewSQL databases
A compromise and fairly young solution type. NewSQL databases are suitable for industries that deal with mission-critical data: financial and banking, healthcare, and others. Companies that care about scalability and transaction reliability.
Benefits:
- easy to scale;
- high performance and data availability.
Disadvantages:
High hardware resource requirements for developers.
Examples: MemSQL, VoltDB, Spanner, etc.
Multidimensional
Multidimensional databases combine several approaches to organizing data into a single server system, so they can store, query, and index it from multiple models. Such databases can convert information from one format to another. They are typically used for IT systems management, data centralization, and big data retrieval.
Advantages:
- consistency of data between models;
- fault tolerance due to ACID compatibility;
- easy integration of new database models into the system;
- suitable for complex projects.
Disadvantages:
- Complexity of multidimensional databases;
- the multidimensional model is still very young and does not have a finalized form;
- limited availability of different modeling methods;
- not suitable for simple projects or systems.
Example: ArangoDB.