On Sep 17, 2010, at 7:30 PM, Nigel Kersten wrote: > On Fri, Sep 17, 2010 at 2:20 PM, Luke Kanies <[email protected]> wrote: > >>> It can be argued that a well laid out external classifier can do this >>> job, but we use environments for our release process, obviously the >>> classifier can't be part of that process, and we have different teams >>> responsible for the classifier (which is part of our puppet >>> infrastructure) and the actual manifests, which are managed by the >>> different platform teams.[1] and [2] >> >> Why can't the classifier be part of that process? > > Just to answer this point by itself, changing the classifier affects > all environments, whereas changing manifests only affects that > environment. > > Someone makes a typo or thinko in the classifier, and suddenly your > production/stable environments are affected. It's not good release > hygiene.
I see. In thinking about it, it's a weird problem - the node classifier is supposed to know all of the nodes that exist, but at the same time you probably want some separation between dev and prod nodes. It should be possible to build environment-like behaviour into the classifiers, but I don't think anyone's done that yet. > It's possible we could set up something like a parent classifier > script that peeked at the client environment in the local server fact > cache and redirected to a per-environment classifier... but frankly > that seems quite hideous. Yeah, and it defeats the whole point of trying to have a consolidated database. -- Luke Kanies http://madstop.com | http://puppetlabs.com | 615-594-8199 -- 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.
