Waratah Documentation

Waratah is an RDF graph database: it stores data as simple subject–predicate–object statements and answers SPARQL queries over them, with built-in support for geospatial and temporal data. It is designed to scale out across many machines; today it runs as a single binary that starts a complete cluster on one machine, which is the shape these pages describe.

Waratah is pre-1.0 software (version 0.3.x). Everything in these pages works today on the single-binary cluster; the multi-machine deployment story is still being built.

Where to start

Getting Started — a cluster running and your first query answered, in about five minutes.
RDF Basics — the five-minute background on statements, IRIs, and literals, if RDF is new to you.
SPARQL Basics — query patterns, worked against the demo dataset.
Command Line — starting, stopping, and maintaining Waratah.
Configuration — the cluster.toml and notebooks.toml files.
Jupyter Notebooks — analyzing your graph from Python.
Geospatial Data — geometry literals, spatial functions, and the S2 index.
Temporal Data — time-typed values, range queries, and the temporal index.

For engine developers

The project overview describes how Waratah is put together, and the developer module carries the full CLI reference, HTTP API contracts, and internals.