When Telco Cloud Automation (TCA) 2.3 is upgraded to TCA 3.0, 3.1 or 3.1.1, after management cluster is upgraded to Kubernetes version 1.26, but before upgrading workload cluster with Kubernetes version 1.24, vsphere-csi addon status on some workload cluster may stuck with Configuring.
3.0, 3.1. 3.1.1
Between Kubernetes version 1.24 in TKG 2.1.1 and TCA 2.3 and Kubernetes version 1.25 in TKG 2.3 and TCA 3.0, the namespace of vsphere-csi addon changed from kube-system to vmware-system-csi. The version of tca-kubecluster-operator running on management cluster with Kubernetes version 1.26 is expecting vsphere-csi addon in vmware-system-csi namespace on the workload cluster. Thus it can not reconcile vsphere-csi addon successfully on workload cluster still with Kubernetes version 1.24.
Resolved in TCA 3.2
Follow this process to upgrade workload cluster of Kubernetes 1.24:
kubectl edit pkgi -n tca-system tca-kubecluster-operatorspec.paused: truekubectl edit deployment -n tca-system tca-kubecluster-operatorspec.replicas from 1 to 0kubectl get po -A|grep kubecOnce workload cluster with Kubernetes version 1.24 is upgraded to version 1.25, vsphere-csi addon status should change to Provisioned.
kubectl edit pkgi -n tca-system tca-kubecluster-operatorspec.paused: truekubectl get po -A|grep kubec