This project provides a devcontainer running PHP that you can use to do your
all of your development work, it has the PHP and PostgreSQL extensions installed
already. There's another container named mysql_db that's running MySQL,
you can use this to persist data for your project. When the containers start
up we also create one running Flyway, it runs the migrations for our
application.
Lastly there's a container named webserver running Apache and PHP, it's
serving the files from the src directory of this project.
If you run on your local machine, you should be able to see the the webserver on port 9080, if you're running on Cloudspace then this port will be forwarded out into it's own URL (in the "Ports" tab of Visual Studio Code). You can connect to the database server using the installed PostgreSQL extension using the default port. The credentials are...
- Server Name: mysql_db
- Account Name: example_account
- Password: applesauce
- Database: example
Note that the database server isn't available outside the container environment, in this project you must use tools from inside Visual Studio Code.
If you need your own migrations, you can open a terminal in your development container and invoke Flyway like so...
flyway migrate
The basic idea here is to provide a semi-complicated development environment so that you can get an idea of how it all works.
You can build and run your devcontainer in the following environments...
- Docker
- Docker Desktop
- GitHub Codespace
It's worth reviewing the first two as it's very easy to get them confused and the abysmal documentation from Docker only makes this worse. The long and short of it is that devcontainers work great on your local machine when your local machine is Linux and decently well on MacOS or Windows as long as your project isn't too big.
If you have a big project and you are on Windows or MacOS then Codespace is the way to go. For my large test project, Visual Studio Code times out during the startup of the container because it's takes so long for the files to make it out to the container. This project will work in all three!
When you are on a Linux machine, you can run Docker on top of your existing Linux installation. In this case Docker is running as a process on your machine, it's just another application you are running. The containers you run are simply more processes on your machine and it all works great.
More importantly, when you share files from your local machine with a container the performance is pretty good.
To connect to the webserver, use http://localhost:9080.
On MacOS or Windows you likely use Docker Desktop to manage your containers. The
way this works is that Docker Desktop creates a virtual machine running Linux,
Docker itself and any containers you may run are processes inside the virtual
Linux machine. Their goofy management application sure is native, the rest of it
very much is not. ;-)
For the most part this works pretty well but when you share files from your local machine what really happens is that those files are shared with the virtual machine that's running Docker. From there, they are shared again with your Docker containers. As you have guessed or even noticed, performance is not great.
To connect to the webserver, use http://localhost:9080.
When you run on Codespace, GitHub provisions Linux containers and handles the sharing of your project with those containers for you. It's all transparent and pretty performant so we're not really concerned with the nitty-gritty of how it works. It works well enough even for large projects.
To connect to the webserver, open the lower panel (where the "Terminal" tab is displayed) and click on the "Ports" tab. There will be one listed and you can click on the URL to open it in your browser.
This project provides an example devcontainer that deploys one development container and a couple other related containers. The development container is on the same network as the other containers, when you connect to the development container with Code you can see everything and do all of your work inside the container. As you make changes you can commit them and push them to GitHub.
This example project is small enough that it will work under MacOS, Windows or Linux. It took me a while to get it all working so I thought I'd write it all down for posterity.
If you notice any bugs or think of a better way to get this done, please send me a pull request!