Plover

Plover is a capistrano script that let's you spin up and shutdown ec2 servers on the fly, and will handle mapping the roles for you. We use it at RailsMachine to spin up staging servers for client code testing. It relies on cloud-init from Ubuntu, the fog gem, and capistrano to do the heavy lifting.

Installing Plover

Plover is a capistrano plugin, so just install it like any other rails plugin and be sure to load the recipe in your rake file.

script/plugin install git@github.com:railsmachine/plover.git

Configuring Plover

Plover uses a simple yaml file (config/plover.yml) to configure options, here is a quick example:

aws_access_key_id: XXXXXXXXXX
aws_secret_access_key: XXXXXXXXXXXXXXXXXXX
groups:
  - my_app
  - ssh
servers:
- flavor_id: m1.small
  image_id: ami-2d4aa444
  role: db
  name: db1.staging.railsmachine.com
  group: db
  options:
    primary: true
- flavor_id: m1.small
  image_id: ami-2d4aa444
  role: db
  name: db2.staging.railsmachine.com
  group: db
  options:
    slave: true
- flavor_id: m1.small
  image_id: ami-2d4aa444
  role: app
  name: app1.staging.railsmachine.com
  groups:
    - app
    - web

Plover will use this configuration to spin up and write out the config/plover_servers.yml file, which contains the server id and dns name of the instance. Plover also uses cloud-init to configure the instance on boot, here is an example cloud-init to run apt-get upgrade and do a custom command:

#cloud-config
apt_upgrade: true
runcmd:
 - [ wget, "http://slashdot.org", -O, /tmp/index.html ]

Learn more about Ubuntu cloud-init

Using Plover

Once it is configured, Plover is very easy to use.

cap plover:provision

and

cap plover:shutdown

License

Plover is released under the MIT license.

---

Unless otherwise specified, all content copyright © 2014, Rails Machine, LLC