Cuprate Docker

Dockerized Cuprate — an alternative Monero node written in Rust

---

Note: Cuprate is under active development and not yet ready for production use. This project is not affiliated with Monero or Cuprate.

Features

• Multi-architecture images (linux/amd64 and linux/arm64)
• Automated builds — new upstream Cuprate releases are detected and built every 8 hours
• Healthcheck via restricted RPC endpoint
• Resource limits and log rotation pre-configured
• Mainnet, testnet, and stagenet support

Quick Start

Docker Compose (recommended)

git clone https://github.com/hundehausen/cuprate-docker.git
cd cuprate-docker
docker compose up -d

Docker Run

git clone https://github.com/hundehausen/cuprate-docker.git
cd cuprate-docker

docker run -d \
  --name cuprate-node \
  -t -i \
  -v cuprate-data:/home/cuprate/.local/share/cuprate \
  -v ./config:/home/cuprate/.config/cuprate \
  -p 18080:18080 \
  -p 18089:18089 \
  ghcr.io/hundehausen/cuprate-docker:latest \
  --config-file /home/cuprate/.config/cuprate/Cuprated.toml

Verify it's running

# Check logs
docker compose logs -f

# Check healthcheck status
docker inspect --format='{{.State.Health.Status}}' cuprate-node

# Query the restricted RPC
curl http://localhost:18089/get_height

Configuration

The default configuration is in config/Cuprated.toml, which is mounted into the container. Edit this file to customize your node. For full documentation, see the Cuprate User Book.

Key defaults:

Setting	Value	Description
network	"Mainnet"	Network to sync (Mainnet, Testnet, Stagenet)
fast_sync	true	Skip verification of old blocks using known hashes
rpc.restricted.enable	true	Restricted (read-only) RPC on 0.0.0.0:18089
rpc.unrestricted.enable	true	Unrestricted RPC on 127.0.0.1:18081 (localhost only)
target_max_memory	auto-detected	Target max memory usage in bytes (auto-detected from system RAM)
target_max_memory	auto-detected	Target max memory usage in bytes (auto-detected from system RAM)

Network Selection

By default the node runs on mainnet. For testnet or stagenet, copy the override example and uncomment the section you need:

cp docker-compose.override.yml.example docker-compose.override.yml
# Edit docker-compose.override.yml, then:
docker compose up -d

Data Persistence

Volume/Mount	Container Path	Description
cuprate-data (Docker volume)	/home/cuprate/.local/share/cuprate	Blockchain database
./config (bind mount)	/home/cuprate/.config/cuprate	Configuration files

Ports

Network	P2P	Restricted RPC
Mainnet	18080	18089
Testnet	28080	28089
Stagenet	38080	38089

Only mainnet ports are mapped by default. See docker-compose.override.yml.example for testnet/stagenet.

Building from Source

# Build with default tag
docker build -t cuprate-docker:local .

# Build a specific Cuprate version
docker build -t cuprate-docker:local \
  --build-arg CUPRATE_TAG=cuprated-0.0.9 \
  .

# Build latest development version
docker build -t cuprate-docker:local \
  --build-arg CUPRATE_TAG=main \
  .

You can also use the compose override to build locally instead of pulling from GHCR — see docker-compose.override.yml.example.

Troubleshooting

STDIN/TTY Issue

cuprated requires a STDIN pipe to avoid log spam (cuprate#396). When using docker run, always include the -t -i flags. The docker-compose configuration handles this automatically with tty: true and stdin_open: true.

Container Exits Immediately

Check the container logs:

docker logs cuprate-node

Common causes: invalid config file syntax, insufficient permissions on mounted volumes. Ensure config/ and its files are readable.

Slow Initial Sync

The initial blockchain sync takes a significant amount of time depending on hardware and network. With fast_sync = true (the default), block verification is accelerated by comparing against known hashes. Monitor progress:

docker compose logs -f

Healthcheck Failing

The healthcheck queries the restricted RPC at http://localhost:18089/get_height. If it fails:

1. Ensure rpc.restricted.enable = true in config/Cuprated.toml
2. The node has a 120-second start period before healthchecks begin
3. Check if the node is still syncing — RPC may not respond until initial startup completes