Closed zer0def closed 2 years ago
@zer0def thanks for updating this! I've merged your changes into my local branch, and am testing with both minikube and gke. I'll report on findings, hopefully that will make it easier for @wunderio to merge.
Hey @snickell, how did your tests go? Generally tested it against upstream K8S, but not sure whether this breaks on peculiar distros, like GKE.
Thank You for creating this. Since configurations might break pre-1.19 clusters, may I suggest moving old yamls to some subfolder (i.e. deploy/kubernetes/1.13
) and putting patched versions in another (i.e. deploy/kubernetes/1.19
)?
Updated per request.
Was finally able to test this on 1.22. There was just one single thing that i'd change, having controller replicas back to 1. I'll change that in a separate commit as I found out installation instructions are also a bit screwed up too (I have not used deploy/kubernetes
resources since initial commit because they were transformed into our helm chart).
Thank You for contribution!
Due to the
storage.k8s.io/v1beta1
API no longer being served by K8S 1.22, it requires a bit of an update to function withstorage.k8s.io/v1
. However, CSIDriver in that API has only been added to K8S 1.19, raising the minimum supported version.