Getting started

Run a NYXDB node with Docker, connect a client, and issue your first append, keyed, streaming, and transform queries.

This section gets a single NYXDB node running and walks through the first queries. It should take a few minutes.

Install & run

Run the engine with Docker or from a local build.

First queries

Append, keyed, streaming, and transform in one session.

The nyxsql CLI

Connect, run one-shot and streaming queries, split panes.

Console tour

Catalog, SQL shell, storage, queries, and lineage in the UI.

NYXDB is in active engine development. The container image and some SQL surface shown here are illustrative of the intended workflow. Commands marked TODO-verify should be confirmed against the release you are running.

The 60-second version

# Run the single-node engine (SQL server on :7777)
docker run --rm -p 7777:7777 -v nyxdb-data:/var/lib/nyxdb \
  ghcr.io/nyxl-io/db:latest

# Connect the CLI (dev default port shown; match your server's --port)
cd drivers/cli && npm install && npm start -- --host 127.0.0.1 --port 7777

Then create a table, insert a row, and read it back — see First queries.