This message was deleted.

Would also really appreciate comments and criticism on my gitops repo layout here, if anyone is around here 🦗

I've been exploring similar challenges with Fleet and don't have all the answers yet, but I wanted to share my current thinking. It seems quite difficult to fully implement a DRY approach with Fleet alone. In light of that, following the WET (Write Everything Twice) approach, as outlined in the Google Cloud documentation, might be worth considering. By allowing some duplication, you gain clarity and explicit control, which can simplify managing configurations across multiple clusters or environments. Additionally, have you thought about using kpt (https://kpt.dev/) to handle rendering manifests before applying them in Fleet? kpt allows for more flexible handling of YAML templates and variables, making it easier to manage the final manifests and the overall process. Here's the Google Cloud guide that introduces the WET repo approach: https://cloud.google.com/kubernetes-engine/enterprise/config-sync/docs/concepts/gitops-best-practices#create-wet-repo Hope this helps!

Thanks for replying @salmon-hair-72590! I gave up on Fleet after giving it a solid try for 2 weeks. The lack of any interaction or community either here or on github, lack of documentation and detailed examples, confusing templating and the need to dig around in the source code to try and figure out what it was doing was too frustrating, I went back to Argo CD 😦

Thanks for the update, I totally understand your decision. I agree, the lack of information, samples, and places to discuss Fleet is definitely frustrating—I've felt the same way. I'm going to stick with Fleet a little longer and see how it goes. Wishing you all the best with Argo CD, hope it works out great for you!