[chef] Re: Re: Access attributes in serverspec tests excuted in a report handler

[Mike < >]

I prefer this method:

>  - Overriding the attributes in .kitchen.yml and hard-coding the

   override values in my tests

This provides a more 'stable' approach of stating with some determinism that when providing some attributes at an input, the cookbook behaves in a known manner.

If you need to test multiple outcomes, add more suites with different attributes.

-M

On Thu, Feb 12, 2015 at 8:00 PM, Phil Mocek < " target="_blank"> > wrote:

"> wrote:
> Is it possible to access the attributes of the current chef run in
> these rspec-tests? And when, how?

Julian C. Dunn responded:
> I don't recommend this approach, because you're conflating the
> variables of the system under test with the test vectors, so how do
> you know if you have a valid test?

What do others recommend doing to verify, for example, that use of a
recipe results in a file existing in a location described in a node
attribute?  The recipe I want to test includes golang::default,
downloads some source code, then creates files in a subdirectory of
node['go']['gopath'].

I'm considering these options:

 - Hard-coding the default values of applicable node attributes
   in my serverspec tests
 - Overriding the attributes in .kitchen.yml and hard-coding the
   override values in my tests
 - [Dumping attributes with a helper cookbook][1], then loading
   the result in my tests
 - Figuring out where the "property" hash in the [apache2
   cookbook's tests][2] comes from and doing something similar

I would appreciate any related suggestions.


References:

 [1]:
<http://jakshi.com/blog/2014/05/12/accessing-chef-attributes-in-serverspec-tests/>
 [2]:
<https://github.com/svanzoest/apache2-cookbook/blob/master/test/integration/default/serverspec/localhost/default_spec.rb>

--
Phil Mocek
https://mocek.org