Why I would want it:
- Easier to do minor tweaks on top of the helm config, if variables exported by helm do not support it
- I find it elegant that all state required to restore state of the cluster are in yaml files. Helm require installation of apps for each cluster
- IIUC if app installed via helm changes between I install it on local cluster and prod cluster I may end up with different versions of the app
I know I can do "kubectl get pod/my-pod -o yaml". Firstly, I would have to figure our which variables are permanent (state variable like "nodeName" are mixed with permanent variables like "restartPolicy"). Secondly, I would also have to figure out all k8s objects that were created by helm.