How are you structuring ArgoCD at scale?

Lately I’ve been thinking a lot about how folks structure their GitOps repositories and workflows with Argo CD once things start growing. In the beginning everything feels simple: a couple services, maybe one cluster, a staging and a prod environment. Almost any structure works. But after some time the platform grows. More services appear, more clusters, more environments, sometimes more teams. At that point the repository structure and the ApplicationSet strategy suddenly become very important

Argocd from scratch, in tree structure etc...

Hello Team, I will start work on a new project, I had experience with argocd for 3,4 years already. In past I used similar setup, but services like nginx-ingress, traefik, csi driver installed with HELM and Terraform in separate infrastructure repo, now I want to utilized ArgoCD for that also. I used ArgoCD, but I feel in past I didn't setup in proper way, some thing we did manually, not via code, (as we used management cluster for Argo, we added managed clusters manually,created argo project

How do you guys run database migrations?

How to implement environments

I am a PA in CS intern, who is tasked with finding the best practices for trying to build a pipeline, that is going to deploy our IaC in the cloud. I have made a basic pipeline which in the CI stage: \- Selects the deployment environment from the branch name (Main = prod, feature/\* hotfix/\* and bugfix/\* = dev, PR = test) \- Validates the IaC and the deployment stage runs the IaC with the various input variables, to the selected Deployment Environment. But my senior engineer has asked m

Dynamic AppSet based on cluster labels

I need a sanity check on what I am trying to accomplish because at this point I am not sure it's doable. I currently have a more complex situation than I have had in past experiences using Argo. I have two on prem clusters and a cloud cluster, with a long list of related services I want to deploy with a single appset. Some services only deploy to one on prem cluster, some to both on prem and some to both on prem and the cloud cluster. I have been trying to deploy to the correct clusters using