Fundamentals and Classification of Databases, CAP Theorem, and Market Trends

Databases are collections of data and logical relationships built, processed, and maintained by specific software known as Database Management Systems (DBMS), which ensure data integrity and security; together they form a Database System (DBS).

Based on data structure, databases can be categorized into traditional relational databases, NoSQL databases, NewSQL databases, and multi‑model databases.

Relational databases, built on the relational model, dominate modern data modeling due to their atomicity, consistency, isolation, and durability; they consist of tables (relations) where each row (tuple) represents a record and each column (attribute/field) defines a data type.

NoSQL (Not Only SQL) databases do not use the relational model, are often distributed, and may not fully adhere to ACID properties; relational databases rely on strong consistency to avoid write and read conflicts.

The CAP theorem states that in the presence of possible network partitions, a system must trade off between availability and consistency; NoSQL databases, which aim for eventual consistency, are well‑suited for high‑load, real‑time scenarios such as e‑commerce and social networks.

NoSQL databases employ various storage models—key‑value, wide‑column, document, graph, as well as extensions like RDF, time‑series, and search engines—to address use cases that relational databases cannot cover.

Due to space limits, only a portion of the content is shared; the full report includes sections on database fundamentals and classification, market status and selection, typical vendor case studies, and future development trends, and is available for download as the "China Database Industry Research Report (2021)".