Here how document-oriented storage works in OrientDB :

Document Structure : In OrientDB, documents are structured as collections of key-value pairs, where each key represents a field or property of the document, and the corresponding value contains the data associated with that field. Documents can have nested structures, allowing for hierarchical representation of complex data.

Schema Flexibility : One of the key features of document-oriented storage in OrientDB is schema flexibility. Unlike traditional relational databases that enforce a rigid schema where tables must adhere to predefined structures, OrientDB allows documents to be stored without a fixed schema or with a flexible schema. This means that documents within the same collection (or class, in OrientDB terminology) can have different sets of fields, and new fields can be added to documents dynamically without requiring schema alterations.

Collections : Documents in OrientDB are organized into collections, which serve as containers for grouping related documents together. Collections can be thought of as analogous to tables in relational databases. Each collection can contain a set of documents, and documents within the same collection share a common structure or class definition.

Indexing and Querying : OrientDB supports indexing of document fields to improve query performance. Indexes can be created on specific fields within documents, allowing for efficient data retrieval based on query predicates. Queries in OrientDB can be formulated using a SQL-like query language called SQL++, which supports querying of document data using familiar SQL syntax.

Atomicity and Durability : OrientDB ensures atomicity and durability of document operations, meaning that individual document updates are transactionally consistent and durable even in the event of system failures or crashes. This ensures data integrity and reliability in document-oriented storage.

Rich Data Types : Document-oriented storage in OrientDB supports a variety of data types for document fields, including primitive types (e.g., string, integer, boolean), complex types (e.g., arrays, embedded documents), and special types (e.g., links to other documents or records).