This message was deleted.

Sorry @best-oil-69865, we'll probably need a bit more to go on then that. Can you provide the logs for the other containers in that pod (e.g. the longhorn-csi-plugin container)? The node-driver-registrar container is trying to connect to a socket created by the longhorn-csi-plugin container. Maybe there is a failure in that one? Also the output of

kubecctl get pod -n longhorn-system -o yaml longhorn-csi-plugin-8p2dh

could be helpful so we can see if there is something nonstandard about how the plugin is deployed.