Watch my Using Docker in Web3 workshop with Developer DAO to see me run through this repo live!
Docker containers can help simplify many of the workflows used in web3. They can be especially useful when you need to run an Ethereum client (such as geth) or network node (such as a Filecoin Lotus node or a Chainlink node), without having to do all of the pre-configuration yourself.
The best option to install Docker for personal use is by downloading Docker Desktop. Docker Desktop is supported on Windows (including support for WSL2), Mac, and Linux. It also has a UI that makes many aspects of building and using containers easier.
Many projects maintain Docker images on Docker Hub. Images hosted on Docker Hub are free to download and install, and are the most common place to find official and community-maintained Docker images.
Many projects will ship a Docker image with their Github repository. The image is created by building a Dockerfile available from a Github repository by following the instructions in the project's README. An example can be seen in the Geth Github repository.
When pulling from Docker Hub, you can pull an image and run a container with it in a single command -
docker run -d --name ethereum-node -p 8545:8545 -p 30303:30303 ethereum/client-go
The docker command acts as our starting point for any Docker operations we'll perform. The run command tells Docker to create a container using the ethereum/client-go image. The -p flag is used to specify what ports on our local machine the container should listen on, allowing us to interact with those ports on the container. The --name flag gives a name to the container so that we can identify it with other Docker commands later. The -d flag runs the container in "detached" mode, which means it will run without connecting our active terminal to the container.
In order to execute a command on our geth node, we use docker exec. One of the most useful ways to use docker exec is to get a live terminal on our container. In order to do that, we can run the command below:
docker exec -it ethereum-node /bin/sh
We can also simply attach to the container, which will let us see the running app's output. In the case of Geth, we should see its connection logs. We can attach with the command below:
docker attach ethereum-node
When a Dockerfile is provided, you must first build the Docker image locally, then run a container with it. Let's use the Geth repo as our example.
git clone git@github.com:ethereum/go-ethereum.git
cd go-ethereum
docker build . -t geth
docker run -d --name test-geth geth
When writing your own software, you can turn it into a Docker image that can be used to easily distribute it to others. For our example, we have a very simple Express server in src/server.ts that we'll use to demonstrate the process.
First, we'll want to build our image:
docker build -t docker-express .
We should see output similar to this:
[+] Building 9.3s (9/9) FINISHED
=> [internal] load .dockerignore 0.0s
=> => transferring context: 2B 0.0s
=> [internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 223B 0.0s
=> [internal] load metadata for docker.io/library/node:16.20 1.6s
=> CACHED [1/4] FROM docker.io/library/node:16.20@sha256:466d0a05ecb1e5b9890960592311fa10c2bc6012fc27dbfd 0.0s
=> [internal] load build context 0.4s
=> => transferring context: 960.98kB 0.4s
=> [2/4] COPY . /opt/express 0.8s
=> [3/4] WORKDIR /opt/express 0.0s
=> [4/4] RUN yarn install 5.6s
=> exporting to image 1.0s
=> => exporting layers 1.0s
=> => writing image sha256:9320eb3a322abe61c58ba19d191944fd2bb05a59da6798dcd6b3fbaad03d2ed8 0.0s
=> => naming to docker.io/library/docker-express
Then, we can create a container using our new image. We'll want to make sure to use the -d flag to run it in the background, and the -p flag to expose port 8080 on localhost. The final command will look like this:
docker run -d -p 8080:8080 --name test-express docker-express
Now that our container is running, we should be able to run a command to test that it's up:
curl localhost:8080
As you can see, our Express server answers the request, just like it would if we had simply run yarn start to start our server!
Now that we've got the basics down, let's talk about some extra tips, and useful features of Docker.
Docker will create resources without names if you want it to, but you should use image tags and container names wherever possible to keep track what you’ve built. This includes using the -t flag on your docker build commands to add tags to your images. The same idea applies to using the --name flag when running your containers with docker run. It's very easy to lose track of what's what when you're stuck trying to identify images and containers with IDs like e7f49ddf3075.
Just like any other software, you want to reduce bloat on your Docker images and containers as much as possible to keep them running quick, and to make sure you don't accidentally reveal something (code, secrets, etc.) to the world through your public images. A helpful tool to handle this is a .dockerignore file.
The .dockerignore file works just like a .gitignore file. Let's say you're using a COPY command in your Dockerfile that gets everything from a directory, but that directory includes a .git folder and a .env file that you don't want to include into the container image. By using a .dockerignore file in the same folder as your Dockerfile, you can prevent those files from making it into your image.
It's possible to mount directories on your local system in a container so that the container has access to them. This makes it possible to avoid accidentally adding your secrets to a Docker image that you'll host publicly on Docker Hub, while still allowing your container to have access to what it needs.
In web3, this is especially useful in situations where your code needs access to a wallet's private key that's stored on your system. In our example, we are storing our .env file in the mnt/ directory. In order to grant our container access to our .env file, we'll run a docker run command that includes the -v flag, as shown below:
docker run -d -p 8080:8080 -v "$(pwd)"/mnt:/opt/express/secret --name mounted-express docker-express
In order to check that our mount is valid, we can use the command below:
docker exec mounted-express cat /opt/express/secret/.env