Skip to content
This repository has been archived by the owner on Dec 19, 2018. It is now read-only.

duggan/pontoon

Repository files navigation

pontoon

A Python CLI (and library) for Digital Ocean.

No Maintenance Intended

Note: This Project is No Longer Maintained

Code and any binaries will remain available.

Consider using the official Digital Ocean CLI tool instead.

Linux CI Windows CI Code Coverage Docs
Build Status on Linux Build status on Windows Coverage Status Documentation Status

Introduction

Pontoon makes interacting with Digital Ocean on the command line smooth sailing.

It is designed for human consumption, and aims to have good defaults.

Caveats

Pontoon has one rule it imposes on top of Digital Ocean:

Names are unique.

Unique names make for a much easier command line experience. For Droplets, names are hostnames, and hostnames should be unique anyway; it's just a good idea.

They don't have to be unique forever though; once a Droplet is destroyed, it's fine to use the name again.

Anything "recommended" (like secure-erasing the drive on termination) is optional, but enabled by default.

Installation

Install via pip:

$ pip install pontoon

To install the latest (development, unstable) release:

$ pip install https://github.com/duggan/pontoon/archive/master.zip

If you'd like to package pontoon for your favourite OS, feel free to do so (and please send a PR to this README!)

More options are on the way.

CLI Usage

Configure

Set up your credentials and preferences:

$ pontoon configure

You'll be prompted for your Digital Ocean API credentials (available here), and whether you want to use existing SSH credentials or for them to be generated (using OpenSSH).

The rest are preferences, and can be set at any time by running configure again, editing the ~/.pontoon config file (YAML format), or by specifying them with options on the command line.

Create your first Droplet!

Now you can create your first droplet:

$ pontoon droplet create foobar
Creating Droplet foobar (512MB using Ubuntu 12.04 x64 in Amsterdam 1)...
.......active

SSH into your Droplet

If everything's configured correctly, you should be able to SSH into your Droplet like so:

$ pontoon droplet ssh foobar
Welcome to Ubuntu 12.04 LTS (GNU/Linux 3.2.0-23-virtual x86_64)

 * Documentation:  https://help.ubuntu.com/
Last login: Fri May  3 18:23:56 2013
root@foobar:~#

List your Droplets

$ pontoon droplet list
foobar:         (512MB, Ubuntu 12.04 x64, Amsterdam 1, 192.0.2.128, active)

or for more detail:

$ pontoon droplet list --detail
foobar
   id:                  998
   name:                foobar
   size:                512MB
   image:               Ubuntu 12.04 x64
   region:              Amsterdam 1
   status:              active
   locked:              False
   private_ip_address:  None
   created_at:          2013-11-09T13:22:40Z
   backups_active:      False
   ip_address:          192.0.2.128

Library

Pontoon bundles a modified fork of python-digitalocean, which can be imported and used as pontoon.lib. The main difference the fork introduces is a system for interacting with mock API responses from DigitalOcean for more thorough testing.

Contributing

Pull requests for bugs are always welcome! New functionality should generally be preceded by a discussion, though if you've written something that you needed and want to contribute back, a pull request is a fine way to start that discussion 🎉

All of the code in pontoon is PEP-8 audited (using pytest-pep8), and there's a full suite of tests written for py.test (library code) and Bats (interface). Contributions should, therefore, include tests and pass a PEP-8 audit.

Running the tests

Tests are run via Tox.

For example, to test the library, CLI and coverage for Python 2.7, run:

$ pip install tox
$ tox -e py27,lib,cli,coverage

The .travis.yml file in this repository enumerates all the tests that are performed.

The CLI tests require BATS, and PEP8 checks are performed in both the lib tests and cli tests.

On OSX, bats can be installed with homebrew:

$ brew install bats

On Debian/Ubuntu, I've set up a PPA for easy installation of bats:

$ add-apt-repository ppa:duggan/bats
$ apt-get update
$ apt-get install bats

The main project is MIT licensed, and the bundled library inherits LGPLv3.

Debugging

Set the DEBUG environment variable (to anything) to enable debug output for pontoon.

This will give a step through of most methods being executed during a command, like so:

$ DEBUG=1 pontoon droplet destroy foobar
2013-11-09 18:37:06,187 [pontoon.configure:DEBUG] combined: (){}
2013-11-09 18:37:06,187 [pontoon.configure:DEBUG] read_config: (){}
Destroying foobar and scrubbing data...
2013-11-09 18:37:06,204 [pontoon.droplet:DEBUG] destroy: (<pontoon.droplet.Droplet instance at 0x10ce1fd40>, 'foobar', False){}
2013-11-09 18:37:06,204 [pontoon.droplet:DEBUG] id_from_name: (<pontoon.droplet.Droplet instance at 0x10ce1fd40>, 'foobar'){}
2013-11-09 18:37:06,204 [pontoon.droplet:DEBUG] list: (<pontoon.droplet.Droplet instance at 0x10ce1fd40>,){}
2013-11-09 18:37:06,205 [pontoon.pontoon:DEBUG] render: (<pontoon.pontoon.Pontoon instance at 0x10ce1fcf8>, 'droplets', '/droplets'){}
2013-11-09 18:37:06,205 [pontoon.pontoon:DEBUG] request: (<pontoon.pontoon.Pontoon instance at 0x10ce1fcf8>, '/droplets'){'params': {}, 'method': 'GET'}
2013-11-09 18:37:07,498 [pontoon.pontoon:DEBUG] render: (<pontoon.pontoon.Pontoon instance at 0x10ce1fcf8>, 'event_id', '/droplets/998/destroy'){'params': {'scrub_data': 1}}
2013-11-09 18:37:07,498 [pontoon.pontoon:DEBUG] request: (<pontoon.pontoon.Pontoon instance at 0x10ce1fcf8>, '/droplets/998/destroy'){'params': {'scrub_data': 1}, 'method': 'GET'}

A timestamp, followed by the module, debug level, the method called and the arguments to that method (positional as brackets, keywords as curlies).

This functionality is implemented by the @debug decorator, the code for which can be seen at pontoon/log.py.

Mocking

Set the MOCK environment variable (to anything) to return mock request data instead of querying Digital Ocean.

This is implemented solely for end-to-end testing of the CLI, but you may find it useful in some other scenarios.