[chef] Re: RE: Re: Re: Re: Problem bootstraping chef-client to Windows machines

[Carlos Camacho < >]

No problem,
Thanks to you for having and maintaining this amazing piece of software!

Actually I'm not a Windows fan but i have to test and use Chef in our
Windows environments, so i will test it for sure.

Regards,

Carlos.




On Tue, Aug 5, 2014 at 2:26 AM, Adam Edwards 
<
 >
 wrote:
> We've released a fix in the Windows cookbook for this issue to Supermarket
> and tested it on different versions of Chef including Chef 10.32.2, Chef
> 11.12.8, and Chef 11.14.2:
> https://supermarket.getchef.com/cookbooks/windows. Please let us know if
> you're still seeing this issue with the new version 1.34.0 of the Windows
> cookbook.
>
> Thank you for your patience, and our apologies for not finding this prior to
> the release of Chef 11.14.2. We are adding additional regression automation
> through test-kitchen to catch this sort of problem on different platforms
> prior to release and even commit.
>
> -Adam
>
> -----Original Message-----
> From: Adam Edwards 
> [mailto:
 
> Sent: Monday, August 4, 2014 8:59 AM
> To: 
> '
 '
> Subject: RE: [chef] Re: Re: Re: Problem bootstraping chef-client to Windows
> machines
>
> Yes, it's definitely a problem in the Windows cookbook, which should
> probably be using chef_gem to bring back the dependency. We'll test out a
> fix to that cookbook and once that's verified get it out to Supermarket
>
> Thank you Carlos, and our apologies for not catching this dependency prior
> to release. The cookbook should not have taken a dependency on Chef
> satisfying its gem dependencies, something we can get better at catching now
> that test-kitchen is getting imrpvoed cookbook testing support on Windows.,
>
> -Adam
>
> -----Original Message-----
> From: Carlos Camacho 
> [mailto:
 
> Sent: Monday, August 4, 2014 7:13 AM
> To: 
> 
 
> Subject: [chef] Re: Re: Re: Problem bootstraping chef-client to Windows
> machines
>
> Ok perfect,
>
> I thought that i was using a chef-client release candidate, maybe its useful
> to only bootstrap new nodes/clients with 1 lower version to mitigate some
> issues and add one more safeguard.
>
> The issue its exactly what you are describing, "Windows cookbook has a
> dependency on ruby-wmi, but doesn't resolve this dependence and it fails"
>
> In this case to have a conclusion, its a problem with the Windows cookbook
> instead the chef-client (Works perfect in Linux as usual)??
>
>
> Best regards,
>
> Carlos.
>
> On Mon, Aug 4, 2014 at 3:43 PM, Steven Danna 
> <
 >
>  wrote:
>> Hi Carlos,
>>
>>>> Its normal to release new client versions with this type of problems!?!?
>>
>> The goal is that such problems are caught in testing and via the
>> distribution of release candidates and betas.  Unfortunately, it seems
>> a problem might have gotten past those safeguards in this case.
>>
>> My suspicion is that the windows cookbook has a dependency on ruby-wmi
>> but doesn't ensure it is installed.  This potentially worked in past
>> versions of Chef because of Chef's own dependencies but changed with
>> the following commit:
>>
>> https://github.com/opscode/chef/commit/e0c2469f3af617c34f13a32cc8b0490
>> 204be7979#diff-ea2858b57306467ffd3011d3ce3e6510
>>
>> I (or someone else) will post some more details once we can look into
>> this problem.
>>
>> Cheers,
>>
>> Steven