On 9 nov 2010, at 11:32, Marc Baudoin wrote:

> Michael Potter <mega...@gmail.com> écrit :
>> On Tue, Nov 9, 2010 at 1:39 AM, Marc Baudoin
>> <baud...@stg-interactive.com> wrote:
>>> 
>>> As the number of different types of hosts in my configuration
>>> grows, I wonder what are the strategies for making bundlesequence
>>> and inputs more modular.
>> 
>> I have dealt with this topic at length. The main problem with a
>> solution to this problem is the inability to extend slists, however
>> there is an alternative, which is to use the ifdefined option to build
>> a list with undefined elements. My current solution would look
>> something like this:
>> 
> [...]
>> 
>>    any::
>> 
>>        "site_seq" slist => {
>>            "@(g.common_seq)",
>>            "@(g.web_seq)",
>>            "@(g.db_seq)",
>>            "@(g.test_seq)",
>>            "@(g.augment1)",
>>        }, policy => "ifdefined";
>> }
>> 
> [...]
>> 
>> Having to add these extra lists each time to the site_seq list is
>> annoying - but in absence of the ability to extend an slist it's the
>> best we have got.
> 
> Sure, better than nothing.  It works for bundlesequence but not
> for inputs.  I just got this error message:
> 
> Can't stat file "/var/cfengine/inputs/@(g.inputs)" for parsing
> 
> Is it possible to use an slist variable with inputs, as in:
> 
> inputs => { @(g.inputs) } ;
> 

body common control
{
ignore_missing_inputs => "true";
}

--
Bas van der Vlies
b...@sara.nl



_______________________________________________
Help-cfengine mailing list
Help-cfengine@cfengine.org
https://cfengine.org/mailman/listinfo/help-cfengine

Reply via email to