Skip to content
This playbook helps you set up and manage your Uberspace(s).
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

Ansible playbook for Uberspace

This playbook helps you set up and manage your Uberspace(s).

It configures a few common things that I find essential for Uberspaces and it is extensible for other stuff.

Current features



  1. Copy uberspaces.example to uberspaces and add your Uberspace host(s) and username(s)
  2. Copy host_vars/ to a new file named without the .example suffix and replace UBERSPACE_NAME with your username, e.g. julia and UBERSPACE_HOST with your Uberspace host, e.g. eridanus.
  3. Add the domains you'd like to run on the respective Uberspace to the file created in step 2.
  4. Repeat steps 2 and 3 for all your Uberspaces.
  5. Run the playbook using ansible-playbook --ask-pass site.yml.
  6. Enjoy!

If you have an SSH keypair and your public key is installed in ~/.ssh/ on your local computer, the key will be stored in ~/.ssh/authorized_keys on your Uberspace and you won't need the --ask-pass argument in subsequent runs.

Let's Encrypt

Nothing to do or configure here. This works automagically for all your domains.


  1. To set up a WordPress instance, simply create an entry under wordpress_instances in your host_vars file (see host_vars/ for an example)
  2. Use the default bedrock_repo from or use your own forked repo of the boilerplate (Your Uberspace's public keys will be conveniently downloaded for you to public_keys/ so you can use them as deploy keys for your private Git repos.)
  3. Add the domains through which your WordPress should be accessible
  4. Make sure to add these domains to the top-level domains section in the host_vars file as well!

Bonus: Deploy hooks

Normally, your WordPress instances will be updated from your repo daily via a cron job. However, if you want to deploy your WordPress whenever your repository changes, you can specify a value for the optional webhook_key in each of your WordPress configs.

With a webhook_key defined, you will be able to create a post-receive hook on your Git server or use your Uberspace as a webhook URL on repository hosting services such as Planio.

Your webhook URLs will be composed like this:

https://{{ uberspace name }}.{{ uberspace host }}{{ wordpress instance name }}.cgi?{{ wordpress instance webhook key }}

A simple post-receive hook on your Git server could look like this, it would have to go in hooks/post-receive:

curl -s ''

Or if you use Planio, simply enter your URL via SettingsRepositoriesyour repoEditPost-Receive webhook URL

Ruby on Rails apps

Setting up and deploying your Ruby on Rails apps involves a little bit more work, but it's definitely worth it. Where else do you get such awesome Rails hosting for the price? Let's get started:

Configure your playbook

  1. To set up a Rails app, create an entry under rails_apps in your host_vars file (see host_vars/ for an example)
  2. Make sure to give it a name which should be only characters, numbers and maybe the underscore character – no spaces!
  3. Add the domains through which your Rails app should be accessible
  4. Make sure to add these domains to the top-level domains section in the host_vars file as well!
  5. For secret_key_base generate a secret using rake secret in your Rails app
  6. For port choose an unused port on your Uberspace between 32000 and 65000

That's it for Ansible. You can now run your playbook using ansible-playbook site.yml.

Configure your Rails app

To actually deploy your app, we're going to use Capistrano. Git clone your Rails app on your local computer and perform the following modifications:

Modify your Gemfile
  1. Add or uncomment gem 'capistrano-rails', group: :development
  2. Run bundle install and then bundle exec cap install
Mofify your Capfile
  1. Add or uncomment require 'capistrano/bundler'
  2. Add or uncomment require 'capistrano/rails/migrations' if your app is using MySQL

Your Capfile should now look similar to this:

require "capistrano/setup"
require "capistrano/deploy"
require 'capistrano/bundler'
require 'capistrano/rails/migrations'

Dir.glob("lib/capistrano/tasks/*.rake").each { |r| import r }
Modify your config/deploy.rb
  1. Find the line set :application, 'my_app_name' and replace my_app_name with the exact same name you chose earlier for the rails_apps entry in your host_vars file.
  2. Find the line set :repo_url, '' and add your Rails app's repo URL. If your repo is private, please add the keys you find in public_keys/ within your playbook as deploy keys to your repository hoster.
  3. Find the line # set :deploy_to, '/var/www/my_app_name', uncomment it and set the value to "~/www/rails/#{fetch :application}" (notice the double quotes!)
  4. Add the line set :linked_files, fetch(:linked_files, []).push('config/database.yml')
  5. Add the line after :publishing, :restart { execute :svc, "-du ~/service/rails-app-#{fetch :application}" } within the namespace :deploy block

Your config/deploy.rb should now look similar to this:

lock '3.5.0'

set :application, 'example_app'
set :repo_url, ''
set :deploy_to, "~/www/rails/#{fetch :application}"
set :linked_files, fetch(:linked_files, []).push('config/database.yml')

namespace :deploy do
  after :publishing, :restart { execute :svc, "-du ~/service/rails-app-#{fetch :application}" }
Modify your config/deploy/production.rb
  1. Find and uncomment the line # server '', user: 'deploy', roles: %w{app db web}, my_property: :my_value and replace with the hostname of your Uberspace, e.g. and deploy with your Uberspace username.

Your config/deploy/production.rb should now look similar to this:

server '', user: 'julia', roles: %w{app db web}

That's it. You should be able to deploy your app using bundle exec cap production deploy. After some time, your Rails app should be humming nicely on your configured domain.


As the Uberspace Playbook is still in development, it would make sense for you to run the cleanup tasks after every update from this repo. The cleanup tasks remove any files/configurations on your Uberspace which previous versions of the playbook may have installed but which are no longer needed. You can run the cleanup tasks like so:

ansible-playbook cleanup.yml




To contribute something you usually configure on your Uberspace, please fork this repo, create a new role (or add to an existing one if it makes sense) and submit a pull request.


I built this. By myself. On my computer.

You can’t perform that action at this time.