Hi everyone! It’s been a while since I’ve been active in the SweetOps community (after participating in the very first office hours with 3 people), but I’ve been following the excellent work Cloud Posse’s been doing over the years for the Terraform community. I’m not sure which channel the best place is to ask this question, but I’m recently been trying to wrap my head around

atmos

and how it can work within companies. There doesn’t seem to be a channel for it yet (and it uses geodesic). There’s two things that I’m a bit unsure about at the moment (with the old reference-architectures in mind): 1. Is the

atmos

approach (with

vendir.yml

) assuming that all modules have the same version (e.g. if you have a

vpc

in multiple workflows) 2. As the reference architectures have been around

atlantis

for automation, I can see some references in the example files for

spacelift

. I’m guessing automation would move away from Atlantis towards Jenkins or other services. Is it still viable to do some form of GitOps with atmos?

Spacelift is gitops and you can use atmos and Atlantis with a bit of craft

@Markus Muehlberger we're slowly releasing the documentation and tutorials

Thanks @PePe Amengual and @Erik Osterman (Cloud Posse). This is tutorial looks really helpful for some things I’m missing. One final question though: Is atmos designed to be a mono repo, or are you still splitting repositories up per AWS account (as I’ve seen different files for dev, staging and prod in the same example)?

We’ve entirely abandoned one-repo-per-account. Mea culpa. That pattern didn’t make gitops easy.

That sounds really amazing! Thanks Erik!