AWS Deploy Helper

A little ruby lib to ease the deployment to an AWS Auto Scaling Group. It connects to an AWS Elastic Load Balancer, see the current instances and generate the ssh and capistrano config files.

Instructions

• Clone this repo in the machine where you are going to execute cap deploy
• Rename config/config.yml.example to config/config.yml and replace all the values
• Run the main.rb script

Result

Assuming you have added two projects (project_one and project_two) to your config.yml, and project_one has 2 instances and project_two has 1 instance right now, you will get these two files:

1. SSH Configuration file

# You will get a file like this at your <<ssh_config_file>> file
Host project_one_1
  Hostname xxxxx1.compute-1.amazonaws.com
  User ubuntu
  IdentityFile ~/.ssh/keys/keys_one.pem  
  StrictHostKeyChecking no

Host project_one_2
  Hostname xxxxx2.compute-1.amazonaws.com
  User ubuntu
  IdentityFile ~/.ssh/keys/keys_one.pem
  StrictHostKeyChecking no

Host project_two_1
  Hostname xxxxx3.compute-1.amazonaws.com
  User ubuntu
  IdentityFile ~/.ssh/keys/keys_two.pem
  StrictHostKeyChecking no  

2. Capistrano Configuration File

# You will get a file like this at your <<cap_config_file>> file
sites:
  project_one:
   - project_one_1
   - project_one_2
  project_two:
   - project_two_1

From now, you can connect to the instances this way:

$ ssh project_one_1
$ ssh project_one_2
$ ssh project_two_1

So, you can read this cap_config_file and proceed to deploy project_one to the instances project_one_1 and project_one_2

Acknowledgment

Thanks to Felipe Espinoza who wrote a preliminar version of the AWSSite Class.