sravan/sravanbalaji.com
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
Files
d103f9224edb431b5dbdaeb76a0987bb0c52826a
sravanbalaji.com/blog_posts/blog_docker.html
Sravan Balaji 8af8ae5180 Directory Rename & Reorganization 
- Rename all directories to follow lowercase and underscore format
- Move to do list from README to GitHub Issues
- Delete BOTW.jpg
- Move common folder out of web pages folder
- Create blog_posts folder and move docker blog post into it
- Fix all page references for new structure
- Move styles.css into common folder
2020-05-27 17:41:40 -04:00

179 lines
14 KiB
HTML
Raw Blame History

<!DOCTYPE html>
<html lang="en">
<head>
<script language="javascript" type="text/javascript" src="../common/header.txt"></script>
<title>Docker</title>
<style>
img {
max-height: 500px;
margin: auto;
display: block;
}
</style>
</head>
<script language="javascript" type="text/javascript" src="../common/navbar.txt"></script>
<body>
<div class="container-fluid px-md-6">
<div class="row align-items-center">
<div class="col-md-2">
<!-- Nothing -->
</div>
<div class="col-md-8">
<!-- Title -->
<h1>Getting Started with Docker</h1>
<!-- Author & Date/Time -->
<p class="lead">
Updated on March 18, 2020
</p>
<!-- Preview Image -->
<img class="img-fluid rounded" src="https://cdn.vox-cdn.com/thumbor/fbrTLtxuP2D29o8VJUaE-u3NKfU=/0x0:792x613/1200x800/filters:focal(300x237:426x363)/cdn.vox-cdn.com/uploads/chorus_image/image/59850273/Docker_logo_011.0.png" alt="">
<hr>
<!-- Post Content -->
<h2><a name="introduction">Introduction</a></h2>
<p>
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, <a href="https://docs.microsoft.com/en-us/windows/wsl/about" target="_blank">Windows Subsystem for Linux</a>, or <a href="https://www.cygwin.com" target="_blank">Cygwin</a>. 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 <b>docker image</b>. You can then spin up an instance of this image called a <b>docker container</b> 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.
</p>
<div class="embed-responsive embed-responsive-16by9">
<iframe class="embed-responsive-item" src="https://www.youtube-nocookie.com/embed/YFl2mCHdv24" allowfullscreen></iframe>
</div>
<hr>
<h2><a name="installing_docker">Installing Docker</a></h2>
<p>
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.
</p>
<ul>
<li>macOS</li>
<ul>
<li><a href="https://docs.docker.com/docker-for-mac/install/" target="_blank">Docker Desktop for Mac</a></li>
<li><a href="https://docs.docker.com/toolbox/toolbox_install_mac/" target="_blank">Docker Toolbox for macOS</a> (Legacy)</li>
</ul>
<li>Windows</li>
<ul>
<li><a href="https://docs.docker.com/docker-for-windows/install/" target="_blank">Docker Desktop for Windows</a></li>
<ul>
<li><a href="https://chocolatey.org/packages/docker-desktop/2.2.0.4" target="_blank">Chocolatey</a>: <code>choco install docker-desktop</code></li>
</ul>
<li><a href="https://docs.docker.com/toolbox/toolbox_install_windows/" target="_blank">Docker Toolbox for Windows</a> (Legacy)</li>
<ul>
<li><a href="https://chocolatey.org/packages/docker-toolbox" target="_blank">Chocolatey</a>: <code>choco install docker-toolbox</code></li>
</ul>
</ul>
<li>Linux</li>
<ul>
<li><a href="https://docs.docker.com/install/linux/docker-ce/centos/" target="_blank">Docker Engine - Community for CentOS</a></li>
<li><a href="https://docs.docker.com/install/linux/docker-ce/debian/" target="_blank">Docker Engine - Community for Debian</a></li>
<li><a href="https://docs.docker.com/install/linux/docker-ce/fedora/" target="_blank">Docker Engine - Community for Fedora</a></li>
<li><a href="https://docs.docker.com/install/linux/docker-ce/ubuntu/" target="_blank">Docker Engine - Community for Ubuntu</a></li>
</ul>
</ul>
<h3>Docker Toolbox & WSL</h3>
<p>
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.
</p>
<h4>Installing WSL</h4>
<p>
The default way to install WSL is through the <a href="https://docs.microsoft.com/en-us/windows/wsl/about" target="_blank">Microsoft Store</a>. 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 <a href="https://kontext.tech/column/tools/308/how-to-install-windows-subsystem-for-linux-on-a-non-c-drive" target="_blank">here</a>.
</p>
<h4>Setting Up Docker Toolbox with WSL</h4>
<p>
After installing WSL, follow the instructions in <a href="https://medium.com/@joaoh82/setting-up-docker-toolbox-for-windows-home-10-and-wsl-to-work-perfectly-2fd34ed41d51" target="_blank">Setting Up Docker Toolbox for Windows Home 10 and WSL to Work Perfectly!</a> starting with <b>Install Docker and Docker Compose within WSL</b>. 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 <code>root = /</code>, you will need to configure shared folders for the <code>default</code> docker machine in VirtualBox.
</p>
<h4>Shared Folder Configuration for Volume Mounting</h4>
<p>
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 <code>docker-machine.exe start</code> to start the <code>default</code> docker machine created by the Docker Quickstart Terminal. After your docker machine is running, open the shared folder settings as shown in <a href="https://medium.com/@Charles_Stover/fixing-volumes-in-docker-toolbox-4ad5ace0e572" target="_blank">Fixing Volumes in Docker Toolbox</a>. The image below shows how <code>C:\</code> on my host computer is mounted at <code>/</code> in folder <code>c</code> on my docker machine such that <code>C:\</code> maps to <code>/c</code>. Make sure you have enabled <code>Auto-mount</code> and <code>Make Permanent</code>. Click <b>OK</b> to save all settings, then go back to the WSL terminal and run <code>docker-machine.exe restart default</code>. Now your docker machine should be properly configured to allow volume mounting.
</p>
<img class="img-fluid mx-auto d-block" src="../assets/blog_posts/Docker/virtual_box_shared_folder.jpg">
<h4>Increase Docker Machine Resources</h4>
<p>
One issue I ran into with the default settings for the <code>default</code> docker machine is the amount of RAM allocated to it. I attempted to build the <a href="https://github.com/borglab/gtsam" target="_blank">GTSAM</a> library inside a docker container and ran into an <i>internal compiler error: Killed (program cc1plus)</i>. As mentioned <a href="https://github.com/docker/for-win/issues/403" target="_blank">here</a>, the solution is to increase the RAM allocated to the docker machine in VirtualBox. In the WSL terminal, run <code>docker-machine.exe stop</code> to shutdown the docker machine. When this is complete, right click on the <code>default</code> docker machine in VirtualBox and open <b>Settings</b>. In the <b>System</b> tab, increase <i>Base Memory</i>. 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.
</p>
<hr>
<h2><a name="building_images_and_running_containers">Building Images & Running Containers</a></h2>
<p>
To learn about Dockerfiles and a basic example, check out the video in the <a href="#introduction">introduction</a>. To learn about Docker Compose, check out Jake Wright's video on the topic.
</p>
<div class="embed-responsive embed-responsive-16by9">
<iframe class="embed-responsive-item" src="https://www.youtube-nocookie.com/embed/Qw9zlE3t8Ko" allowfullscreen></iframe>
</div>
<hr>
<h2><a name="x_forwarding">X Forwarding for GUI Applications</a></h2>
<p>
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 <a href="https://sourceforge.net/projects/vcxsrv/" target="_blank">VcXsrv Windows X Server</a>. If you are using <a href="https://chocolatey.org/packages/vcxsrv" target="_blank">chocolatey</a>, you can run <code>choco install vcxsrv</code>.
</p>
<p>
<a href="https://dev.to/darksmile92/run-gui-app-in-linux-docker-container-on-windows-host-4kde" target="_blank">Run GUI app in linux docker container on windows host</a> 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 <code>DISPLAY</code> environment variable that is passed to docker via the <code>-e DISPLAY=$DISPLAY</code> command line argument. You can alternatively pass your IP address as an environment variable in your docker compose file. This is shown in <a href="https://github.com/MAAV-Software/ros-tutorial" target="_blank">MAAV's ROS Tutorial</a> (reproduced below). The main settings to look at below are <code>environment</code> and <code>network_mode</code>. Set the <code>DISPLAY</code> environment variable to <code>[ip_address]:0.0</code> and set <code>network_mode: "host"</code>.
</p>
<pre>
<code>
# 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
</code>
</pre>
<hr>
<h2><a name="helpful_commands">Helpful Docker Commands</a></h2>
<p>
<ul>
<li>System Prune to clear space on Docker Machine</li>
<ul>
<li><code>docker system prune --volumes</code></li>
</ul>
</ul>
</p>
<hr>
<h2><a name="conclusion">Conclusion</a></h2>
<p>
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 <a href="../web_pages/about.html#contact_me">contact me</a>.
</p>
<hr>
</div>
<div class="col-md-2">
<!-- Nothing -->
</div>
</div>
</div>
</body>
<footer>
<script language="javascript" type="text/javascript" src="../common/footer.txt"></script>
</footer>
</html>
Reference in New Issue View Git Blame Copy Permalink