datafiles

Datafiles are computer files that store data to be read and manipulated by software. They may contain structured data such as database records or unstructured data such as text, images, or multimedia. Datafiles serve as the persistent storage for applications and are located on storage devices; they can include raw data and, in many systems, metadata, indexing, or checksums to support data integrity and access.

In database management systems, datafiles are the physical containers of user data. Oracle uses datafiles as the components of tablespaces; a tablespace may comprise one or more datafiles, and data is stored in fixed-size blocks within them. SQL Server differentiates data files (.mdf for primary data, .ndf for secondary data) from separate transaction logs (.ldf). PostgreSQL stores its data in a data directory as files managed by the database engine, with tables, indexes, and other objects mapped to on-disk files. Other systems use similar concepts with varying terminology.

Datafiles are not limited to database storage. They include common interchange formats such as CSV, JSON, XML, and Parquet, as well as binary formats used by applications or machine learning pipelines. They may be plain text or binary, and often employ compression, encryption, or schema information in headers or accompanying metadata to aid interoperability and security. File size, growth policies, backups, and integrity checks are important considerations for datafile management. In databases, redundancy and replication of datafiles, together with logs, help ensure durability and recoverability.

See also: database file, file format, data storage.