Run a Taiko node with Docker

This guide will help you start up a Taiko RPC node using simple-taiko-node.

Software Dependencies

Dependency	Version	Version Check Command
git	^2	git --version
Docker	^24.0	docker --version

Prerequisites

• If using Windows, you should install Git BASH or WSL to use as your terminal.
• Meet the Geth minimum hardware requirements except for the storage requirement because Taiko nodes will require less storage (at the time of writing).

Run a Taiko node with simple-taiko-node

1. Clone simple-taiko-node

   Mac/Linux

   git clone https://github.com/taikoxyz/simple-taiko-node.git
   cd simple-taiko-node

   Windows

   git clone https://github.com/taikoxyz/simple-taiko-node.git
   cd simple-taiko-node && git config core.autocrlf false

2. Copy the sample .env files

   Info

   To run a hekla Taiko node please copy .env.sample.hekla instead of .env.sample.

   cp .env.sample .env

3. Set the L1 node endpoint

   First, open the .env in your preferred text editor:

   Mac/Linux

   nano .env

   Windows

   notepad .env

   Note

   You can use any L1 endpoint to run a Taiko L1 node unless you want to prove blocks past the last 128 blocks, in which case you will need an archive node.

   It’s recommended to run a local L1 node but you browse around for other RPC Providers. Keep in mind they will eventually rate limit your node and it will stop syncing, so a local L1 node is recommended for a proper setup.

   For Mainnet, the L1 Node will need to be an Ethereum node; for Testnet, the L1 Node will need to be a Etherem testnet (Holesky) node.

   Next, you will set the L1 node endpoints. If you are running a local L1 node, you cannot reference the L1 endpoints as http://127.0.0.1:8545, ws://127.0.0.1:8546 and http://127.0.0.1:5052 because that is local to inside the simple-taiko-node Docker networking. Instead you can try:

   • Using host.docker.internal (see: stack overflow).
   • Using the private ip address of your machine (use something like ip addr show to get this ip address).

   After getting the address of the L1 node, set the following L1 node endpoints in your .env file. Here is an example:

   • L1_ENDPOINT_HTTP=http://192.168.1.15:8545
   • L1_ENDPOINT_WS=ws://192.168.1.15:8546
   • L1_BEACON_HTTP=http://192.168.1.15:5052

4. Remove old testnet volumes

   If you ran a testnet node previously, make sure to first remove the old volumes:

   docker compose down -v

5. Set the profiles you want to run

   In your .env, please set COMPOSE_PROFILES to whichever set of services you’d like to run in a comma separated list (i.e., l2_execution_engine,proposer,prover or any combination of the three).

   Note that if you include proposer or prover, the services will still be disabled if you do not set ENABLE_PROPOSER=true or ENABLE_PROVER=true respectively.

6. Start the node

   Info

   Make sure Docker is running and then run the following command to start the node (you may need to use sudo docker compose up -d if your user is not in the docker group):

   To run a hekla Taiko node please use the following command instead: docker compose -f docker-compose-hekla.yml up -d.

   docker compose up -d

7. Verify node is running

   Option 1: Check with the node dashboard

   A node dashboard will be running on localhost on the GRAFANA_PORT you set in your .env file, which defaults to 3001: http://localhost:3001/d/L2ExecutionEngine/l2-execution-engine-overview.

   You can verify that your node is syncing by checking that the chain head on the dashboard is increasing. Once the chain head matches what’s on the block explorer, you are fully synced.

   Option 2: Check with curl commands

   1. Check if the Execution Layer client is connected to Taiko L2:

      curl http://localhost:8547 \
        -X POST \
        -H "Content-Type: application/json" \
        --data '{"method":"eth_chainId","params":[],"id":1,"jsonrpc":"2.0"}'

      which should return the chainId as 0x28c61 (167009):

      { "jsonrpc": "2.0", "id": 1, "result": "0x28c61" }

   2. Check if the Execution Layer client is synced by requesting the latest Taiko L2 / L3 block from the Execution Layer client:

      curl http://localhost:8547 \
        -X POST \
        -H "Content-Type: application/json" \
        --data '{"method":"eth_blockNumber","params":[],"id":1,"jsonrpc":"2.0"}'

   3. If the blockNumber response value is 0 or not growing, check the Taiko L2 logs here:

      docker compose logs -f

      Note: You may need to use sudo docker compose logs -f if you are not in the docker group.

Video tutorial

See the video tutorial Run a Taiko L2 node (YouTube). It’s for a previous testnet but the overall steps are the same!

Full simple-taiko-node CLI reference

Make sure you are in the simple-taiko-node folder. If you are not:

cd simple-taiko-node

Start node

docker compose up -d

Stop node

docker compose down

Restart node

docker compose down && docker compose up -d

Update node

git pull origin main && docker compose pull

Remove node

docker compose down -v

Start and remove orphan containers

This command will start and remove containers that are not part of the current docker compose file.

docker compose up -d --remove-orphans

View grafana dashboard

open http://localhost:3001/d/L2ExecutionEngine/l2-execution-engine-overview

View logs

docker compose logs -f

View execution logs

docker compose logs -f l2_execution_engine

View client driver logs

docker compose logs -f taiko_client_driver

View client proposer logs

docker compose logs -f taiko_client_proposer

View system resource usage stats

docker stats

Compare .env and .env.sample

sdiff .env .env.sample

Troubleshooting

• Visit the Discord for help on any common error codes / issues.
• View the logs of the node (see above).