> I'm not very familiar with ruby, so feedback is welcome. This patch is just > for review and must definitely be improved. I would first of all like to > know if this is the way things should be done or if I've completely missed > the point.
Thanks for the patch, it's great to have something to discuss and think about the implications. If you're ever on irc feel free to ping me (nasrat). > My idea was to be able to pass a "ttl" parameter to any fact, cache the > result in a file and read the value from this file if it's not newer than > the TTL. It would be used this way: > > # cache fact for one hour > Facter.add(:myfact, :ttl => 3600) do > [...] > > I mainly had in mind the need to avoid constantly probing web services > which aren't updated that frequently. But maybe Luke had something else in > mind when reporting the issue ? > > Before continuing, I would appreciate some input on the following > questions: > > - is it sufficient to just store the value in plain text ? Or should it be > serialized in yaml or whatever ? For now we're just key/value, we will be working on a model that is richer, so having a cachable interface and being able to switch the backend is the way to go. > - would it be an idea to have infinite ttl (:ttl => -1) ? Presuming some > fact values would never changes. I think so, > - where and how should facts get cached ? Obviously /tmp/<factname> was > just for testing. Ah just read this after commenting on the patch! Paul --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
