Getting Started with Docker

Updated on March 18, 2020

Introduction

Throughout my college career in computer science, I have found the need to develop applications that will be deployed in a Linux environment. As a windows user, I have usually turned to a couple of different options: virtual machines, Windows Subsystem for Linux, or Cygwin. Recently, I discovered an alternative method for building and testing Linux apps that is especially useful for collaboration and deployment. Docker allows you to package an OS and all the necessary dependencies into a docker image. You can then spin up an instance of this image called a docker container in which you can run your source code. Jake Wright's video was incredibly helpful in my understanding of what docker is, how it's different from virtual machines, and how to use it.

Installing Docker

Depending on your host machine's operating system, you will install one of the following Docker products. If you have macOS or Windows, your first option should be Docker Desktop. Check if you meet the system requirements. If you do not, Docker Toolbox is the legacy option that uses Oracle VirtualBox rather than Hyper-V or HyperKit for virtualization. I am using a computer with Windows 10 Home, so I use Docker Toolbox for Windows. This blog post will discuss some solutions to edge-case issues with Docker Toolbox that Docker Desktop users will hopefully not have to deal with.

Docker Toolbox & WSL

As mentioned previously, I am using Docker Toolbox since Windows 10 Home does not support Hyper-V for virtualization with Docker Desktop. You can run docker toolbox in a command prompt or powershell terminal, but I personally prefer to use WSL because I am more familiar with bash terminal commands. If you are interested in this setup, continue reading this section. Otherwise, you can skip ahead.

Installing WSL

The default way to install WSL is through the Microsoft Store. This will install WSL on your system drive (e.g. C:\). If you are like me and don't want to install it on your system drive, follow the instructions here.

Setting Up Docker Toolbox with WSL

After installing WSL, follow the instructions in Setting Up Docker Toolbox for Windows Home 10 and WSL to Work Perfectly! starting with Install Docker and Docker Compose within WSL. As you will have seen near the end of that post, there are some extra steps needed to ensure that volume mounts work. After setting root = /, you will need to configure shared folders for the default docker machine in VirtualBox.

Shared Folder Configuration for Volume Mounting

I configured all of my drives as shared folders so I can mount any folders on my host machine to any docker containers I run. In your WSL terminal, run docker-machine.exe start to start the default docker machine created by the Docker Quickstart Terminal. After your docker machine is running, open the shared folder settings as shown in Fixing Volumes in Docker Toolbox. The image below shows how C:\ on my host computer is mounted at / in folder c on my docker machine such that C:\ maps to /c. Make sure you have enabled Auto-mount and Make Permanent. Click OK to save all settings, then go back to the WSL terminal and run docker-machine.exe restart default. Now your docker machine should be properly configured to allow volume mounting.

Increase Docker Machine Resources

One issue I ran into with the default settings for the default docker machine is the amount of RAM allocated to it. I attempted to build the GTSAM library inside a docker container and ran into an internal compiler error: Killed (program cc1plus). As mentioned here, the solution is to increase the RAM allocated to the docker machine in VirtualBox. In the WSL terminal, run docker-machine.exe stop to shutdown the docker machine. When this is complete, right click on the default docker machine in VirtualBox and open Settings. In the System tab, increase Base Memory. Similar to virtual machines, I would not recommend allocating more than half of your system's RAM. For example, my computer has 16 GB of RAM, so I would only recommend allocating up to 8 GB of RAM to the docker machine. You can look through the other options, such as number of allocated processor cores, as well if you wish.

Building Images & Running Containers

To learn about Dockerfiles and a basic example, check out the video in the introduction. To learn about Docker Compose, check out Jake Wright's video on the topic.

X Forwarding for GUI Applications

In a virtual machine, you can easily run GUI applications such as a web browser, text editor, etc. In a docker container, the process is a little bit different. You will need to download an X Server such as VcXsrv Windows X Server. If you are using chocolatey, you can run choco install vcxsrv.

Run GUI app in linux docker container on windows host provides a great explanation of why you would want to do this and how to run your docker container. The post discusses how to set your IP address as the DISPLAY environment variable that is passed to docker via the -e DISPLAY=$DISPLAY command line argument. You can alternatively pass your IP address as an environment variable in your docker compose file. This is shown in MAAV's ROS Tutorial (reproduced below). The main settings to look at below are environment and network_mode. Set the DISPLAY environment variable to [ip_address]:0.0 and set network_mode: "host". 

                        
# docker-compose.yml
version: "2"

services:
ros-demo:
    image: ros-demo
    privileged: true
    volumes:
    # Mount the current directory do everything in /tutorial within docker
    - .:/tutorial:rw
    environment:
    - DISPLAY=1.1.1.1:0.0 # Change 1.1.1.1 to your IP Address
    network_mode: "host"
    container_name: ros-demo
    command: "/bin/bash --init-file scripts/source-ros.sh" #source ros automatically

Helpful Docker Commands

  • System Prune to clear space on Docker Machine
    • docker system prune --volumes

Conclusion

Hopefully by this point, you have a better understanding of what docker is, why you might want to use it, and how to get a docker machine setup so you can build and run docker containers. If you have any questions or suggestions for improvement, feel free to contact me.