nwaku-compose

Ready‑to‑use docker compose stack for running your own nwaku node.

📝 Prerequisites

• Docker and Git
• Linea Sepolia RPC endpoint You can get a free endpoint from Infura or any other Linea Sepolia RPC provider.

🚀 Starting your node

#	Setup Type	Setup Style	What happens
A	manual	Power User	Setup a .env file manually and then start the node.
B	setup-wizard	Command-Line	Generates .env, starts the node.

🧪 OPTION A :- manual [ recommended ]

1. Setup .env file

cp .env.example .env  

Edit the .env file and fill in all required parameters

💽 2. Set Database Parameters

Waku uses PostgreSQL to store and serve messages.
Limit disk usage and (optionally) increase shared memory for better performance.

Setting	Auto-set command	Manual example
Storage size	./set_storage_retention.sh	echo "STORAGE_SIZE=50GB" >> .env
Shared memory	./set_postgres_shm.sh	echo "POSTGRES_SHM=4g" >> .env

🖥️ 3. Start your node

Start all processes: nwaku node, database and grafana for metrics.

docker compose up -d

⚙️ OPTION B :- setup-wizard [ experimental ]

Run the wizard script. Once the script is done, the node will be started for you, so there is nothing else to do.

The script is experimental, feedback and pull requests are welcome.

./setup_wizard.sh

📌 Note

RLN membership is your access key to The Waku Network. It is registered on-chain, enabling your nwaku node to send messages in a decentralized and privacy-preserving way while adhering to rate limits. Messages exceeding the rate limit will not be relayed by other peers.

If you just want to relay traffic (not publish), you don't need to perform the registration.

---

Monitor your nwaku node

• Metrics (Grafana):
  Open localhost:3000 to view node metrics.

• Live logs:
  See what’s happening inside your node in real time:

  docker compose logs nwaku -f --tail 100

• Use the REST API

  Your nwaku node exposes a REST API to interact with it.

  # get nwaku version
  curl http://127.0.0.1:8645/debug/v1/version
  # get nwaku info
  curl http://127.0.0.1:8645/debug/v1/info
  

For advanced documentation, refer to ADVANCED.md.

How to update to latest version

We regularly announce new available versions in our Discord server.

From v0.35.1 or older

Please review the latest https://github.com/waku-org/nwaku-compose/blob/master/.env.example env var template file and update your .env accordingly.

Also, move your Sepolia RPC client (e.g., Infura) to a Linea Sepolia RPC client.

You will need to delete both the keystore and rln_tree folders, and register your membership again before using the new version by running the following commands:

1. cd nwaku-compose ( go into the root's repository folder )
2. docker-compose down
3. sudo rm -r keystore rln_tree
4. git pull origin master
5. ./register_rln.sh
6. docker-compose up -d

From v0.36.0 or newer

Updating the node is as simple as running the following:

1. cd nwaku-compose ( go into the root's repository folder )
2. docker-compose down
3. git pull origin master
4. docker-compose up -d

Node's health check

Once done, check your node is healthy:

./chkhealth.sh 

All good:

02:15:51 - node health status is:

{
  "nodeHealth": "Ready",
  "protocolsHealth": [
    {
      "Rln Relay": "Ready"
    }
    ...
  ]
}

If the ./chkhealth.sh script is hanging or returns the following, wait a few minutes and run it again:

02:17:57 - node health status is:

{
  "nodeHealth": "Initializing",
  "protocolsHealth": []
}

Disk cleanup tips

Docker artefact can take some precious disk space, run the following commands to free space while your node is running.

Only do this if this machine is solely used for Waku and you have no other docker services.

I repeat, this will clean other docker services and images not running, only do this if this machine is only used for Waku.

# Be sure that your containers **are running**
sudo docker-compose up -d

# Clean docker system files
sudo docker system prune -a

# Delete docker images
sudo docker image prune -a

# Delete docker containers
sudo docker container prune

# Delete docker volumes
sudo docker volume prune

journal

If your /var/log gets quite large:

journalctl --disk-usage
> Archived and active journals take up 1.5G in the file system.

You can cap the size in /etc/systemd/journald.conf with

SystemMaxUse=50M

then restart to apply

systemctl restart systemd-journald

and verify

journalctl --disk-usage
> Archived and active journals take up 55.8M in the file system.

FAQ

[see](FAQ.md)