On 04/07/12 13:42, DEGREMONT Aurelien wrote: > Le 04/07/2012 13:06, Brice Figureau a écrit : >> >> Well, you could try those on your (test) cluster and report back what it >> gave (only the master needs to be actually updated) :) > We did and it works fine, but, as you said, this is our test cluster and it > is not so representative. > And all your explanations about how all of this was done in your specific > cases makes me feel shy to test it on real > production environment :)
Yes, I can understand that. You might be able to change only one of your master (if you have more than one of course), and point only some of your nodes on it, then stay like this for a while, until adding more nodes if things still work. You can also use RI tools to compare catalogs[2] and gain confidence about the output... > So I was waiting for feedback from Puppet Labs guys to be sure. What I know for sure, is that the patch seems to work, but I really didn't test it enough to be confident there are no side-effects. Since the patch is small, it is quite easy to understand it. The only way I can see it fail is to prevent some manifests changes to be picked-up by the master. > By the way, I can't wait to see your patches to also speed up the puppet > agent ;) That's a whole different topic :) Fixing #2198 [1] would be a very good start, then parallelize non-dependent sub-trees. In a word, that's not easy. [1]: https://projects.puppetlabs.com/issues/2198 [2]: https://github.com/ripienaar/puppet-catalog-diff -- Brice Figureau My Blog: http://www.masterzen.fr/ -- 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.
