Standalone/Headless mode without ArgoCD dependency #1385
-
|
First of all, congratulations on version 1.0. I manage multiple ArgoCD instances across different clusters, and some of them don't have internet access. So I would like a centralized image updater that:
Currently, Image Updater requires ArgoCD Application resources to function, even when using Git write-back mode. This makes centralized management across multiple ArgoCD instances impractical. I know this is mentioned as one of your limitations, but I'm wondering if you would be willing to remove that limitation? Proposed SolutionA "headless" mode where image tracking configuration is defined independently (via CRD or config), without requiring ArgoCD Applications. The updater would:
Challenges I Understand
What would this solve?This would also solve the problem of cronjobs or other applications that also scales to 0 like Keda would start to get image update without having to wait for ArgoCD to report which version that is currently running. What are your thoughts around this? |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments
-
|
Thanks for considering argocd-image-updater in your work. What you're describing is an interesting challenge, but does not align well with current argocd-image-updater design. Our future plan will continue to focus on improving the current CRD-based architecture and configuration to better support workflow centered around argocd. |
Beta Was this translation helpful? Give feedback.
-
|
That is what I expected. Thanks for your quick answer. |
Beta Was this translation helpful? Give feedback.
Thanks for considering argocd-image-updater in your work. What you're describing is an interesting challenge, but does not align well with current argocd-image-updater design. Our future plan will continue to focus on improving the current CRD-based architecture and configuration to better support workflow centered around argocd.