On Thu, Dec 9, 2010 at 23:04, Trevor Vaughan <[email protected]> wrote:
> Thanks for the response.
>
> Would it not be possible to keep the entire structure in memory until
> writing the final file?
Not with the current architecture of puppet; this is one of the long
term issues I personally want to see fixed. Right now, there is no
"final" available to the provider, and no way to indicate which parts
belong together, so it is a non-trivial undertaking (as far as I can
see) to implement.
> Would a possible solution to this be a puppet-concat daemon that the
> Puppet types communicate with over a named pipe (or whatever)? How much
> benefit would this add?
It would be possible, but I suspect the complexity of implementation
would outweigh the benefits. You might do better thinking about how a
better algorithm for change detection might be implemented. (I can't,
off hand, identify one, but it probably isn't impossible. An in-core
provider /might/ also be able to achieve that without some of the
costs - less forks, at least.)
> Trying to scope the work involved before I potentially try to implement it.
Moving concat into Ruby? Not too bad, probably 12 hours I would
guess, including all the tests. More than that? Very hard.
Regards,
Daniel
--
✣ Daniel Pittman ✉ [email protected] ☎ +61 401 155 707
♽ made with 100 percent post-consumer electrons
--
You received this message because you are subscribed to the Google Groups
"Puppet Developers" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/puppet-dev?hl=en.
