Binary Installation
Run CoordiNode directly on your machine — no Docker required.
Download a Release Binary
Pre-built binaries for Linux and macOS are available on the GitHub Releases page.
# Linux x86_64
curl -L https://github.com/structured-world/coordinode/releases/latest/download/coordinode-linux-x86_64.tar.gz \
| tar -xz
sudo mv coordinode /usr/local/bin/
# macOS (Apple Silicon)
curl -L https://github.com/structured-world/coordinode/releases/latest/download/coordinode-macos-arm64.tar.gz \
| tar -xz
sudo mv coordinode /usr/local/bin/Verify:
coordinode --versionBuild from Source
Requirements: Rust 1.80+, protoc 3.21+
git clone https://github.com/structured-world/coordinode.git
cd coordinode
cargo build --release -p coordinode-serverThe binary is at target/release/coordinode.
Start the Server
coordinode serve --data /var/lib/coordinodeKey flags:
| Flag | Default | Description |
|---|---|---|
--addr | [::]:7080 | gRPC listen address |
--ops-addr | [::]:7084 | Health + metrics endpoint |
--data | ./data | Data directory |
--peers | (none) | Peer addresses for cluster mode |
See Configuration for the full flag and environment reference, the packaged config file, and the open-file-descriptor limit.
The binary exposes gRPC only on port 7080. For REST/JSON, run structured-proxy in front of it (see docker-compose.yml for a reference setup).
Verify
curl http://localhost:7084/health
# → {"status":"serving"}Compact After a Bulk Import
After loading a large dataset, run an offline compaction to collapse the accumulated merge operands (adjacency lists, counters) into single values. This keeps traversal reads fast. The server must be stopped first.
coordinode compact --data /var/lib/coordinodesystemd Service (Linux)
[Unit]
Description=CoordiNode graph database
After=network.target
[Service]
ExecStart=/usr/local/bin/coordinode serve --data /var/lib/coordinode
Restart=on-failure
User=coordinode
[Install]
WantedBy=multi-user.targetsudo systemctl daemon-reload
sudo systemctl enable --now coordinodeNext Step
See Quick Start to seed data and run your first hybrid query.
