Self-Hosting a Syndicate Appchain Node

Run your own Syndicate appchain RPC node with Docker Compose.

Prerequisites

• Docker with the Compose plugin (v2.20+)
• curl and tar (for snapshot download)
• Disk space: 1 TB+ recommended
• RAM: 16 GB+ recommended
• CPU: 4+ cores recommended
• Ports 8545 (HTTP RPC) and 8548 (WebSocket RPC) available on the host
• WebSocket RPC access to your sequencing chain and settlement chain (e.g. via Alchemy, Infura, or a self-hosted node)

Note

For synd-TEE withdrawals to be functional, you'll need to set up a synd-enclave on an AWS Nitro TEE instance and provide that URL as ENCLAVE_RPC_URL. See the TEE enclave setup guide for instructions.

Setup

1. Copy the env template

cp .env.example .env

2. Fill in .env

Open .env and supply your chain-specific values. All initial values for your specific appchain can be provided by the Syndicate team.

Warning

BATCHER_PRIVATE_KEY and PROPOSER_PRIVATE_KEY must have funds on the sequencing / settlement chains respectively. Additionally, the BATCHER must be authorized to sequence on the sequencing contract.

To change the location where data is persisted, set DATA_DIR in your .env before running start.sh.

3. Start

bash start.sh

The script will:

1. Create local data directories under DATA_DIR (default: ./data)
2. Download and extract the nitro snapshot if NITRO_SNAPSHOT_URL is set
3. Start all services with docker compose (the mchain container handles its own snapshot download via MCHAIN_SNAPSHOT_URL if set)

Verify

Check that all containers are running:

docker compose ps

Wait for the mchain and nitro to finish syncing, then test the RPC:

curl -s -X POST http://localhost:8545 \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","method":"eth_blockNumber","params":[],"id":1}'

WebSocket RPC is also available at ws://localhost:8548.

Common commands

# Follow logs for all services
docker compose logs -f

# Follow logs for a specific service
docker compose logs -f nitro

# Stop all services
docker compose down

# Stop and remove all persisted data (bind-mounted)
docker compose down && rm -rf ${DATA_DIR:-./data}

# Restart a single service
docker compose restart mchain

Data directory layout

./data/
├── mchain/       # RocksDB state for the intermediate chain node
├── nitro/        # Nitro node state
├── sequencing/   # filesystem cache for the sequencing ingestor
├── settlement/   # filesystem cache for the settlement ingestor
└── valkey/       # Valkey (Redis-compatible) persistence

Summary

At this point you should have a functional synd-stack rollup deriving state from the parent chains. You can assert that the rollup node is synced by checking the eth_blockNumber rpc call result. You should also be able to send new transactions by calling eth_sendRawTransaction on the rollup node. Withdrawals should also be functional, you can assert this by monitoring the TEEModule contract for assertionPosted and closeChallengeWindow events (example)