anaconda cookbook

Chef cookbook for installing Continuum Analytic's
Anaconda: "completely free Python
distribution for large-scale data processing, predictive analytics, and
scientific computing".

This also serves as a live example of the most up-to-date best practices for
writing, maintaining, and testing Chef cookbooks:

• Berkshelf 3 for dependency resolution
• Vagrant for development
• Chefspec for rapid testing
• Test Kitchen for comprehensive testing across multiple platforms, with tests written in serverspec
• Foodcritic for style checking

Requirements

This repo has only been tested with RVM; YMMV with other installation methods
(rbenv, chef-dk, etc).

• Berkshelf 3.1.3
• Chefspec 4.0.0
• Test Kitchen 1.2.1
• Foodcritic 4.0.0
• Vagrant 1.6+
  • vagrant-omnibus
  • vagrant-berkshelf: note that >= 2.0.1 is required

It sounds like Chef-DK is the new
recommended installation path, but I have not had a good experience with it (as
of 0.1.0-1). Again, YMMV.

Quickstart

The [Vagrantfile](Vagrantfile) is written to get you an Anaconda environment
with minimal effort:

$> vagrant up --provision
...

$> vagrant ssh
$vagrant> export PATH=/opt/anaconda/2.0.1/bin:${PATH}
$vagrant> conda --version
conda 3.5.5

# if you included `recipe[anaconda::shell_conveniences]` you don't have to do anything;
# it's sourced in /etc/profile.d
$> vagrant ssh
$vagrant> conda --version
conda 3.5.5

To use it in a cookbook:

include_recipe 'anaconda::default'

Usage, recipes, attributes, and resources

The main recipe is anaconda::default. Include it in your runlist, and it will
install the package as well as any necessary dependencies.

The following are user-configurable attributes. Check
[attributes/default.rb](attributes/default.rb) for default values.

• anaconda
  • version: the version to install
  • flavor: either x86 (32-bit) or x86_64 (64-bit)
  • install_root: the parent directory of all anaconda installs. note that individual installs go into #{install_root}/#{version}
  • accept_license: must be explicitly set to the string yes; any other value will reject the license.
  • owner: the user who owns the install
  • group: the group who owns the install

recipe[anaconda::shell_conveniences]

Include this to have the environment set for all users (login shells) via
/etc/profile.d. Useful for development.

$> vagrant ssh
$vagrant> which conda
/opt/anaconda/2.0.1/bin/conda

resource anaconda_package

You can use the anaconda_package resource to install new packages into the
Anaconda environment:

# I do not know what 'astroid' is, just using it as a sample package
anaconda_package 'astroid' do
  # the other supported action is `:remove`
  action :install
end

See the [resource definition](resources/package.rb) for additional options; in
general, all it does is present the same options as conda install/conda
remove.

Tests

Run the full test suite:

# this will take a really long time
$> script/cibuild
...

# check the final result; bash return codes: 0 is good, anything else is not
$> echo $?

Run just the [chefspecs](spec/default_spec.rb):

$> rspec

Run just the test kitchen serverspec [integration
tests](test/integration/default/serverspec/default_spec.rb):

# this is what takes so long: every platform and version is fully built in vagrant
# the list of OSes is defined in [.kitchen.yml](.kitchen.yml)
$> kitchen verify

# test a specific OS; `kitchen list`
$> kitchen verify default-ubuntu-1204

Check the style with Foodcritic:

$> foodcritic

Releases and issues

Standard stuff:

• master is the active version in development
• releases are made with Github, and git tag'ed

Issues should be opened in the Github issue
tracker

TODO

• autodetect 64-bit versus 32
• (TODO does it matter? who uses it?) populate metadata.rb: suggests, supports, etc
• add a pre-provision for kitchen tests to avoid redownloading the installer on every test (really slows down the tests)
• figure out how to publish onto http://community.opscode.com/; the documentation is unbelievably bad

Author

Author:: Matt Chu (matt.chu@gmail.com)