I fixed this by replacing the csi-cinder-controllerplugin and the csi-nodeplugin using the manifests found in this project:
https://github.com/kubernetes/cloud-provider-openstack.
I used kubectl to make the changes. Does anyone know if end-users can configure these kinds of changes when kolla-ansible installs magnum? Or when creating the cluster template?
Hello,
When I try to attach a persistent cinder volume to a pod, I get FailedMount and FailedAttachVolume timeout events. I also get these errors in the log of the csi-cinder-controllerplugin-0 pod:
E1020 13:38:41.747511 1 reflector.go:126] k8s.io/client-go/informers/factory.go:133: Failed to list *v1beta1.VolumeAttachment: the server could not find the requested resource
E1020 13:38:41.748187 1 reflector.go:126] k8s.io/client-go/informers/factory.go:133: Failed to list *v1beta1.CSINode: the server could not find the requested resource
I fixed a CrashLoopBackoff error with the csi-snapshotter container in the csi-cinder-controllerplugin-0
pod by providing the label "csi_snapshotter_tag=v4.0.0" when I created the cluster template. I found that suggestion in an issue on the GitHub cloud-provider-openstack project.
I'm not finding
any help with this error on Google.
Thanks,
-Stu