- From: Christopher Webber <
>
- To:
- Subject: [chef] Re: cookbook verification
- Date: Tue, 12 Aug 2014 07:59:45 -0700
So this is one of those huge yak shaves that really depends on what you are
doing.
When you say macro and micro levels I think less in terms of single cookbook
vs multiple cookbook and more in terms of single node vs multi node.
With that distinction in mind the first thing I would do is take a gander at
a presentation @someara did at Chef-NYC:
http://vimeo.com/98938732.
For me the workflow is:
- Linting tools (used to make the code a bit easier to deal with and clean)
- Foodcritic
- Rubocop
- Unit or Pre-converge testing (Really fast feedback. Tests that I told Chef
to do the right thing)
- ChefSpec
- Integration/Functional or Post-converge testing (Did I automate the right
thing? Was the automation effective.)
- Test Kitchen
- ServerSpec
From a multi-node level, I really haven’t had a chance to wrap my head around
this yet. But in an ideal world, you would have a test or staging environment
and you could push your Chef changes there, then run ServerSpec or whatever
monitoring you already run in prod against that environment.
That is my 2cents anyway.
— cwebber
On Aug 11, 2014, at 4:28 PM, Stuart Cracraft
<
>
wrote:
>
>
Hello!
>
>
What are the best methods for
>
doing cookbook verifications at
>
the micro and macro levels?
>
>
I.e. the unitary single cookbook
>
itself, alone and the greater universe of all it's peer cookbooks.
>
>
--Stuart
>
Archive powered by MHonArc 2.6.16.