This repository has been archived by the owner on Jul 15, 2024. It is now read-only.
Replies: 1 comment
-
I could potentially publish something to k8s (such as configmap with the application/tag information) and use the Cluster Decision Generator for that. https://argocd-applicationset.readthedocs.io/en/stable/Generators-Cluster-Decision-Resource/ |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
We are designing our git strategy such that an Application has 2 repos, 1 for the application code itself, with the "contract/interface/abstraction" being that repo "exports" a Docker image. It does not care how it is deployed or to where, so as such it does not include k8s manifests.
We then have a sibling repo for the manifest and other related infrastructure for that application.
This allows us to be very clear about changes. Changes to config/infra do not result in a new application build.
My question is, with this type of architecture, I'm trying to figure out the best way to "marry" the application version (the docker image tag), with that of the infrastructure code.
ArgoCD may be pointing at the infrastructure repo (to get the k8s manifest templates), but I need something like the
git generator
, or even maybe adocker generator
? something that allows me to identify an image (tag) that can be passed into the templates.I've looked at things like https://github.com/argoproj-labs/argocd-image-updater, which feels like what I want, but it only supports helm/kustomize and we will likely not be using either (possibly using Tanka/Jsonnet or CDk8s).
https://github.com/keel-hq/keel also operates too late in the deployment pipeline. I don't want to decouple the image update process from the deployment of manifests itself.
Beta Was this translation helpful? Give feedback.
All reactions