but both the local management cluster and a rancher launched RKE cluster both have the issue, but both also work just fine from the UI