vagrant-docker

---

PROJECT DESCRIPTION

This project helps you manage your docker containers with a small footprint.

---

Personal requirements

It is built on top of Virtualbox and Vagrant, therefore you should understand what both does, before you try to use this tool here. Also you should known how to create a Dockerfile.

---

INSTALLATION

• Install Virtualbox

• Install Vagrant (Homebrew variant was not tested)

• Mac User: You need to install at least the command line tools. See more here.

• Open your command line and run the following commands to install all required plugins (Terminal on Mac OS X/ CMD.exe on Windows)

• vagrant plugin install vagrant-hostmanager vagrant-vbguest nugrant

• Download this project into your local machine and rename the unzipped file to vagrant-docker

• Copy projects.txt.dist to projects.txt

• (Optional) Copy .vagrantuser.dist to .vagrantuser. Change the values for your needs. (Ex. the vm-name, IP)

• Open your command line and navigate to the vagrant-docker directory.

• Run the command vagrant up and drink a coffee or something.

• Run the command vagrant provisionto update your local host file.

• Open your browser and enter phpmyadmin.dev.

• In case you do not see phpMyAdmin, something simalar like this demopage, then you forget one step or something has changed in the meantime.

---

Add your own Project

In this project you find a directory called dockerfiles/. This is the place where you store your dockerfiles in the following structure:

dockerfiles/
dockerfiles/private/
dockerfiles/private/myservice/run.sh
dockerfiles/private/myproject1/run.sh
dockerfiles/private/myproject2/run.sh
dockerfiles/company/
dockerfiles/company/companyservice/run.sh
dockerfiles/company/companyproject1/run.sh
dockerfiles/company/companyproject2/run.sh

After vagrant runs for the first time a directory will be filled with services from https://github.com/reinblau/dockerfiles.

Every project/service needs at least one shell-script called run.sh. This file is used to start your project (container) while vagrant starts the shell provisioner. This means when you run the command vagrant up or vagrant reload or simple vagrant provision.

To detect an error the script should exit with a none-0 exit code in case of an error. To reuse another service from dockerfiles/ you can call rbrequire --project=directoryname before you start your container. Take a look at https://github.com/reinblau/dockerfiles/tree/master/drupal_boilerplate to see an example of it. As you’ll see we use crane to build images or start a container. You can do that as well!

---

Updates

The shell provisioner provides a update mechanism to update this project, the OS, all the dockerfiles/ which contains a git-repository, and all projects which have a crane.yml file. The container will then be stopped and removed to start the new container from the fresh docker image.

• By default the shell provisioner will not update the system etc automatically, todo this change the value of autoupdate to true in your .vagrantuser

• To update the system manually: Log in to the VM via vagrant ssh and run the command rbupdate

To avoid the update of the dockerfiles or a docker image, create a file blacklist.txt and enter the directory names of every "namespace" or project. Example:

public
custom
myproject

• Add your own update script (update.sh) per git repository or per project, which will be called after the update mechanism run git pull or run crane provision for a enabled project. This is usefully to pull docker image or build new docker image. Example: dockerfiles/company/update.sh or dockerfiles/company/myproject/update.sh

---

Hosts file

Our shell provisioner reads the VIRTUAL_HOST environment variable (provided for the jwilder/nginx-proxy container) from all projects and creates a file called vhosts.txt. The vagrant plugin vagrant-hostmanager use this text file and updates your hosts file.