Volume Creation fails with Duplicate name error about ibm-powervs-block-csi-driver

[Madhan-SWE]

The controller will always retries to create volume again and again. The request will always fail with Duplicate Name error.
Thus, the workloads will always stuck in Pending state.

[root@madhan-multinode-kubernetes-1 ~]# kubectl get pods --all-namespaces | grep Pending
kube-system-test   app-2-11                                                0/1     Pending   0          65m
kube-system-test   app-2-23                                                0/1     Pending   0          65m
kube-system-test   app-2-25                                                0/1     Pending   0          65m
kube-system-test   app-2-3                                                 0/1     Pending   0          65m
kube-system-test   app-2-31                                                0/1     Pending   0          65m
kube-system-test   app-2-41                                                0/1     Pending   0          65m
kube-system-test   app-2-46                                                0/1     Pending   0          65m
kube-system-test   app-2-47                                                0/1     Pending   0          65m
kube-system-test   app-2-49                                                0/1     Pending   0          65m
kube-system-test   app-2-53                                                0/1     Pending   0          65m
kube-system-test   app-2-55                                                0/1     Pending   0          65m
kube-system-test   app-2-58                                                0/1     Pending   0          65m
kube-system-test   app-2-59                                                0/1     Pending   0          65m
kube-system-test   app-2-7                                                 0/1     Pending   0          65m

[mkumatag]

The controller will always retries to create volume again and again. The request will always fail with Duplicate Name error.
Thus, the workloads will always stuck in Pending state.

not sure if any other providers got a logic of checking the volume presence before create calls to avoid these errors?!

[Madhan-SWE]

There was a method to repair existing volume in Ceph-csi
Reference: https://github.com/ceph/ceph-csi/blob/30333378ef39eaf6dd53d06434752db798f56ba1/internal/rbd/controllerserver.go#L370

Need to explore other csi drivers and find an efficient way of handling this.