Draft:ArcadeDB

ArcadeDB is an open-source multi-model database supporting document model, graph model, and key-value model. Due to its non-relational nature, this is a NoSQL database. ArcadeDB runs on the Java platform and is written in Java, particularly only using Java's low-level API for performance reasons. Originally, this project started as a fork of OrientDB. The development is driven by the community conducted on the github platform. ArcadeDB is listed in the DB engines ranking and in Carnegie Mellon's Database of Databases.

Multi-Model

 * Key-Value Model: A record is a pair of a unique key and a data value.
 * Document Model: A record is a sets of key-value pairs.
 * Graph Model: A record is either a vertex document or an edge document.

where:


 * Types are an optional, partial schema of a document, vertex or edge.
 * Buckets are containers for documents, vertices or edges of a certain type.

Features

 * Schema-full, Schema-less, and hybrid/mixed mode
 * ACID compliant
 * Interfaces: HTTP/REST, Java, JDBC, MQL, Redis, PostgreSQL,
 * Supported data query languages: SQL (dialect), Cypher, Gremlin , GraphQL, MQL
 * Importable formats: OrientDB, Neo4j, XML, JSON, CSV, RDF
 * High availability via replication
 * Consistency via raft consensus algorithm
 * Container support via Docker / Podman and Kubernetes
 * Studio web user interface

Native Query Language
The native query language of ArcadeDB is a dialect of SQL, which is closely related to OrientDB's OSQL, and extends a subset SQL with functionality for documents, vertices, edges, buckets, types, and properties.

Programming Language Support (Drivers)
ArcadeDB natively includes support for Java. The community provides support for .NET, Python, Ruby, Rust, and Chicken Scheme. Generally, the HTTP API endpoint interface allows connectors, drivers, or clients in most programming languages.

Usage
ArcadeDB provides a terminal client (console) for interacting with the database server and also a web interface (studio) for database management tasks, as well as queries and commands in all supported query languages and inspect query results in tabular text, raw JSON, or Graph rendering.

Testing
ArcadeDB has been tested in the context of scientific investigations of graph databases in and .

Planned Functionality
A time series model and a vector embedding database model are currently under construction.

History

 * 2021-09-01 Initial Release of ArcadeDB.

Name
The name ArcadeDB refers to the underlying business model of only R&D and word of mouth, and no dedicated sales team.