Good point, I think some efforts were done in field but evidently they 
might not be enough.
As for the opportunity to extend to defines the data binding functionality 
I definitively agree, now, that it could be a severe performance issue so I 
wonder if it might make sense to consider a sort of meta-parameter that 
would enable such functionality on demand: disabled by default but 
activable by users or authors.
Such a parameter might be applied also to classes.
If not possible or advisable on a per define/class basis it could make 
sense as a configuration option, so that who doesn't use data bindings can 
disable it. 

Just guessing...
Al

On Monday, May 19, 2014 6:47:41 PM UTC+2, Joshua Hoblitt wrote:
>
> It seems to me that there are two reoccurring basic themes in this 
> thread of folks expressing either: 
>
> * eagerness to push more data into hiera 
>     or 
> * sentiment that hiera is already a performance bottleneck 
>
> The first is language / design pattern issue while the later is an 
> engineering problem.  Has much effort been spent looking into 
> accelerating hiera's resolution speed? 
>
> -Josh 
>
> -- 
>

-- 
You received this message because you are subscribed to the Google Groups 
"Puppet Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/puppet-dev/986a73d8-44ec-429a-9ece-2a7c754e5a32%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to