On My Travel - Infrastructure

The first version of the application has been set-up on an Ubuntu server hosted on a virtual server.

As a proof of concept, it was enough to install the dependencies manually. But, as we try to make the application a bit more professional, we built this repository in order to automate the application deployment.

** Disclaimer ** : This reposimarktory intents to be as up-to-date as possible BUT it is set to work on an Ubuntu server 14.04x64. Some incompatibilities may or may not occurs.

Requirements

• Vagrant
• Ansible

Set-up

First, open a terminal and go into the repository clone folder. Then, start the virtual machine by executing the following commands.

cd vagrant/
vagrant up

It should take a moment, especially if you have never downloaded the Ubuntu package that is required. When the process is completed, let's check that the SSH connection to the local virtual machine is allowed to ansible.

cd ../ansible/
ansible all -m ping -i inventories/setup -u vagrant --ask-pass

Where all is the group of machine we want to ping (all in this case) and vagrant is the ssh username.

It should ask you to enter a password. Type vagrant (default password for the default vagrant user) and the ansible command should answer:

127.0.0.1 | SUCCESS => {
    "changed": false,
    "ping": "pong"
}

First step

The first step to set the vagrant machine ready to host the application is to start with this first playbook. It creates the ansible user and secure the SSH connection.

But first, it's necessary to generate an SSH key that you will use to authenticate yourself on the server. This step is optional if you already have an id_rsa keyfile.

As usual, to generate your SSH key:

ssh-keygen -t rsa -b 4096 -C "<YOUR EMAIL HERE>"

It will ask you to select the directory where you want to generate the SSH key. It is usually created as /Users/<YOU>/.ssh/id_rsa. You can either use this one or change the SSH keypath in the project.

Then, it's simple:

ansible-playbook -i inventories/setup setup.yml

From now, you should be able to connect on the virtual machine using the following command. The IP is set in the vagrant file to 172.28.128.3 and the default user is ansible. If you changed things, do not hesitate to update it accordingly.

ssh ansible@172.28.128.3 -p 2222

If everything goes OK, then you can run the next command. Obviously, it is important to pick the right inventory (local for vagrant, master for the server).

Install the dependencies

The next step is to secure the server and install the few dependencies (including the database) that are required for the application to run.

This command should not been used very often, only when the application infrastructure changes.

ansible-playbook -i inventories/local install.yml

Deploy the application

This one is supposed to be called every time you need to deploy a application version. It will fetch the new master version on the Github repository and install it on the server.

ansible-playbook -i inventories/local deploy.yml

Access localy to the application

If you want to run the application localy, and access the API and the webfront client within the virtual machine running with vagrant, you would have to modify you /etc/hosts/ file, adding the next line.

172.28.128.3 dev-onmy.travel api.dev-onmy.travel