Getting started

This section takes you from "what is RushDB?" to a running node with data in it. Read it in order if you are new to RushDB; each page is short and ends with links to the deeper material.

RushDB is a standalone, persistent property-graph database that runs as one process. You write JSON records, connect them with typed relationships, and query them through a native JSON protocol or through SQL, GQL, openCypher, Gremlin, SQL/PGQ and Mongo-style statements.

Pages

1. Introduction — what RushDB is, the data model in one paragraph, the query surfaces, and what is deliberately not in the engine.
2. Installation — install with one command, Homebrew, Docker, Cargo or a release archive; start a node with rushdb up; choose a data directory; and set your own access token.
3. Quickstart — a ten-minute tour: create records, ingest nested JSON, query, link records, run a vector search, and connect with psql.
4. Concepts — short definitions of record, label, property, relationship, tenant, transaction, snapshot, index and capability profile, each linking to its full page.

Where to go next

• Data model — how RushDB stores records, properties, relationships and tenants.
• Querying — the native query document in depth.
• Connectivity — HTTP, PostgreSQL wire, gRPC, Bolt and the CLI.
• Operations — deploying and running RushDB in production.