Leverage the power of docker to make development initialization easy. The goal is to have the members of your team ready to code in three steps:

1. Installing Docker
2. Installing Docker Compose
3. Cloning this repo
4. Create a symlink to the repo in their clone of the application.
5. Running docker-compose up

Step 1 through 3 will only be needed once in the life of their machine. Step 3. and 4. can become optional if you have an image registry. This would leave only… step 5. That's right, just one command.

For the moment there is :

• an nginx image
• a php image
• a postgres image
• a memcached image
• a mailcatcher image
• an elasticsearch image

All images are based on the same docker image : ubuntu:trusty.

The project contains a sample Symfony2 application. You can do the same thing in your app.

1. Create a docker directory, in order to have a clean separation from the rest of the project.
2. Create a Dockerfile. It will contain the definition for the data container of your application. It should consist only of VOLUME statements, defining which files and directories will be shared with the containers. Make sure to only use VOLUMES corresponding to containers you will be using.
3. Create any configuration file needed by the containers you intend to use. Configuration files could be part of the image rather than mounted on it, that is true, but having them as part of your application is great, because it makes explicit what will be needed when deploying the application.
4. Create a symlink to the clone of this project, it will be used to tell docker-compose where to look for images. It should not be under version control.
5. Create a docker-compose.yml file and map the volume with paths on your host.

By default, defines a monitoring.dev virtualhost. Try it, it displays nginx and php-fpm status pages.

It comes with XDebug remote debugging ready to use! XDebug will try to reach your DBGP client on 172.17.42.1:9000, please ensure that your client is not only listening on localhost. You will also need to configure remote path mappings in your debugging client so that the remote /srv folder is mapped to where your application is on your system.

If you are using vim with vdebug, the configuration will look like this:

let g:vdebug_options= {
\    "server": '172.17.42.1',
\    "path_maps": {"/srv": "/home/username/path/to/code"}
\}

The recommended way to login to this container is to exec the login command on it, like this:

docker exec --interactive --tty samplesymfonyapp_appserver_1 /bin/login -p -f $(whoami)

The -p flag tells login to preserve the environment, which contains SSH_AUTH_SOCK, used when authenticating against remove service that provide composer packages. The -f flag will allow us to bypass authentication.

Exposes an administration interface on port 80, and an SMTP service listening on port 25.

To use the DNS functionality, you have to run 'bin/docker-init' with root privileges

sudo bin/docker-init

In your docker-compose.yml file you can set DNSDOCK_NAME and DNSDOCK_IMAGE for each image that you want to subscribe in DNS. Additionaly, you can set the DNSDOCK_ALIAS to completely bypass the basic naming scheme

webserver:
    build: ./docker/docker-images/nginx
    ports:
        - '80'
    environment:
        - DNSDOCK_IMAGE=web
        - DNSDOCK_NAME=my_project
        - DNSDOCK_ALIAS=my_app.dev

With that configuration you can access to your image with my_project.web.docker and my_app.dev

No need to add any host in your hosts file.