not only SQL (NoSQL)

NoSQL (Not-only SQL) is a collective term for databases without a relational structure. The term was first used in 1998.

NoSQL databases, many of which are opensource projects, enable non-relational data storage with flexible table formats or with graphical node structures. Since additional servers can be added by scaling out, they offer particularly high scalability. In addition to inexpensive processing of large amounts of data, the user also receives a high level of fail-safety and system availability

Variants of NoSQL databases include key-value databases, large-column databases, document-oriented and graph databases

Key-value databases store data under a key. Although highly scalable, these databases offer limited search capabilities. In contrast, in-memory computing( IMC) provides high computational speed by keeping data in memory all the time, while on-disc versions store data on a data store before processing.

Large-column databases have multiple tables with many columns and scale highly, but no relationships are established between tables.

Document-oriented databases such as Apache CouchDB store documents in Javascript Object Notation( JSON), but without a data schema and in an arbitrary structure. Entries of arbitrary length can be accessed based on the document contents.

Graph databases such as Neo4j map and provide access to data as graphs rather than tables. While an SQL database can only calculate the relationship between different elements with great effort, graph databases represent the stored data as nodes as well as the relationships between these nodes.

While relational databases handle multiple tables and data sets across them, NoSQL solutions primarily offer high scalability. NoSQL does not mean a departure from the relational data model of Structured Query Language (SQL), but rather a supplement depending on the application environment. In this context, NoSQL systems are recommended for the flexible handling of variable data, for mapping relationships or for reducing a database to a container for the key values.