Hi >> @@ -68,6 +69,9 @@ class Facter::Util::Resolution >> @confines = [] >> @value = nil >> @timeout = 0 >> + if :ttl >> + @cache_file = "/tmp/" + name.to_s >> + end > > You probably should be requiring tempfile and using that as we do > support windows and it'll do the right thing. Also we'd probably want > to ensure we don't have race conditions here.
as far as I understood tempfile removes the created file when the object is finalized (if not told earlier with unlink), so after each run, the files are gone. aren't they? > I'd probably think we'd be better served with a cachable interface and > a tempfile implementation like this as being one particular > implementation (eg puppet could use an in memory cache), but this is a > great start. A common interface would be great to address the different problems on different platforms. One issue with tempfiles we should keep in mind is that the different distributions/platforms are handling tmp-folders differently. And for example on RedHat-based Systems tmp-folders are imho the wrong location as tmpwatch is managing them and introducing another cache manager ;) I already had various problems with applications writing to /tmp or /var/tmp and assuming that the files remain there for the whole applications lifetime. So on *nix systems I think /var/cache/facter would maybe the place to go? No idea for windows. Otherwise + 1 for the patch. Thanks a lot! cheers pete --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To post to this group, send email to puppet-dev@googlegroups.com To unsubscribe from this group, send email to puppet-dev+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/puppet-dev?hl=en -~----------~----~----~----~------~----~------~--~---
