Setup a PV bound to an existing file share. Then a PVC bound to that PV.
Need to update the capacity for both PV and PVC.
PV is updated OK, but PVC does not update the capacity properly:
Warning VolumeResizeFailed 34s (x8 over 2m8s) external-resizer file.csi.azure.com resize vol │
│ ume "pvc-sandboxes-files-agudinya" by resizer "file.csi.azure.com" failed: rpc error: code = Inval │
│ idArgument desc = GetFileShareInfo(account-name_file-share-name) failed with error: error parsing │
│ volume id: "account-name_file-share-name", should at least contain two #
The capacity request for the PVC is properly updated, but the actual available capacity not.
The only way to update now is to recreate the PVC.
Only left alternative is to create a whole bunch of new PV+PVC with the new capacity, reasign workloads, and delete old PV+PVC.
What you expected to happen:
Not sure what the behavior should be when BYO Share is used in terms of capacity management, I did not find anything in the docs about it. The most natural behaviour I believe would be that in these cases no capacity management at all is done by the driver.
What happened:
Setup a PV bound to an existing file share. Then a PVC bound to that PV.
Need to update the capacity for both PV and PVC.
PV is updated OK, but PVC does not update the capacity properly:
The capacity request for the PVC is properly updated, but the actual available capacity not.
The only way to update now is to recreate the PVC.
Only left alternative is to create a whole bunch of new PV+PVC with the new capacity, reasign workloads, and delete old PV+PVC.
What you expected to happen:
Not sure what the behavior should be when BYO Share is used in terms of capacity management, I did not find anything in the docs about it. The most natural behaviour I believe would be that in these cases no capacity management at all is done by the driver.
Environment:
Kubernetes: 1.29.4 CSI driver: 1.30.2