Hi Julien, On Sun, 2020-09-06 at 16:10 +0200, Julien Lecomte wrote: That sounds interesting. I don't have any experience with Prometheus, but it would definitely fit into the DebOps role collection. Be aware though that the apache role is not well maintained. You may be better off using the nginx role instead. Your role is most welcome if it is well-written, well-documented and in any way useful for DebOps users. It would also be nice if you can maintain it for a while, which means that you should be actively using your role yourself. Reviewing is done by one or more other DebOps contributors and the role is ultimately approved or denied by Maciej (@drybjed). My experience is that the reviewing process of new roles can take a while. We're trying to streamline that. Be aware of the other two open (and well-discussed) PRs for Prometheus-related roles: https://github.com/debops/debops/pulls?q=is%3Apr+prometheus+is%3Aopen Cheers, Imre

Hey there, Imre pretty much covered it, I just wanted to chime in on a couple of things: I'm in the process of writing a "Writing roles" doc that will hopefully make the standards for role inclusion explicit and simplify the whole endeavor. Unfortunately I'm moving house and won't be able to be too hands-on for a week, maybe a bit more. :( For now, new additions found in the changelog may be considered the ones following the most current style, maybe have a look at those for inspiration. Using facts provided by core debops roles to populate defaults is generally encouraged. fhs is one of those, off the top of my head. The Universal Configuration guide has a few pointers on how the end-user config is expected to behave, utilizing the ansible_plugins role. Also, ideally we hope role authors agree to be listed as codeowners for their roles in order to have a hand in future maintenance. I'm not an expert on Prometheus by any stretch, but #1310 looks pretty much ready, perhaps helping review and test that will be all that's needed. Tasos