Brief Tutorial on Docker

Writing the Dockerfile

Docker can build images automatically by reading the instructions from a Dockerfile, which is a text file that contains all commands needed to build a container image. We show an example of Dockerfile in the following.

FROM ubuntu:20.04
ARG DEBIAN_FRONTEND=noninteractive

# Install software packages inside the container
RUN apt-get update  \
    && apt-get -y install  \
          iputils-ping \
          iproute2  \
          net-tools \
          dnsutils  \
          mtr-tiny  \
          nano      \
    && apt-get clean

# Put file inside the container
COPY file  /

# The command executed by the container after startup
CMD [ "/bin/bash"]

The above Dockerfile indicates that our container is being built on top of the official Ubuntu 20.04 Docker image. That is the purpose of the FROM command. On top of this base image, we use the RUN command to install a number of additional software packages, including some network utilities and the nano editor. The "apt-get clean" command at the end clears the local repository of retrieved package files that are left in /var/cache, so the final image size can be reduced. More explanation can be found in the comments.

Building and starting the container

We will statically assign IP addresses to our container. To do that, we need to attach our container to a network first. We can let docker attach our container to the default network. If you can take this option, you can skip the following network-creation step.

In many situations, we would like to attach our containers to a custom network. To achieve that, we need to first create a network using the following docker command, which creates a network called seednetwork, and its network prefix is 172.20.0.0/24. We can further use "docker network" command to manage all the created networks (listing, deleting, etc.).

$ docker network create --subnet=172.20.0.0/24 seednetwork

We are now ready to build and run our container. The first command in the following builds a container called seedhost using the Dockerfile in the current directory. The second command starts the container. The flags -it tell Docker we want an interactive session with a tty attached. The --rm option will remove the container when it stops. The --ip option specifies the static IP address, and the --net option specifies the network this container should be attached to.

$ docker build -t seedhost .
$ docker run --net seednetwork --ip 172.20.0.5 --rm -it seedhost

Some related commands. Docker has many useful commands, we refer students to online resources (there are many) to learn those commands. In the following, we summarize some of the commands commonly used in SEED labs.

• docker ps: List all the running containers, including the ID for each container. The container id is a unique alphanumeric number for each container, and it is needed in many commands.

• docker ps -a: List all the containers, regardless of whether they are running or not.

• docker exec -it <container id> /bin/bash: Execute the /bin/bash command in a running container. We will get a shell prompt. You don't need to type the entire ID string; typing the first few characters will be sufficient, as long as they are unique.

• docker rm <container id>: Remove a container.