pv is getting stuck in terminating state after upgrading the external provisioner to v5.1.0 latest release

[ramandeepsharma]

What happened:

pv is getting stuck in terminating state after upgrading the external provisioner to v5.1.0 latest release. The issue seems to be caused by patch permissions which got added here: PR-1155. We have to delete the finalizer as well as pv manually currently.

What you expected to happen:

pv shouldn't get stuck in terminating and should get deleted without removing the finalizer.

How to reproduce it:

Anything else we need to know?:

We also noticed the following driver logs during pv termination and looks like because of these errors, provisioner is not being able to delete the PV.

Warning VolumeFailedDelete 44s (x379 over 23h) efs.csi.aws.com_efs-csi-controller-6d9d89f7c7-xxxx rpc error: code = NotFound desc = Failed to find access point for volume: fs-xxxxxx

Environment:

• Driver version: v2.0.7 (helm chart version: 3.0.8)
• Kubernetes version (use kubectl version): v1.29
• OS (e.g. from /etc/os-release):
• Kernel (e.g. uname -a):
• Install tools:
• Others:

[Jainbrt]

@ramandeepsharma what is exact minor version of K8s, you are running ?

[ramandeepsharma]

@Jainbrt Exact K8s version is 1.29.8

[Jainbrt]

it should work with 1.29.8, while I am hoping that you are using the latest provisioner too (5.1.0)

[ramandeepsharma]

Yes I tried it with 1.29.8 and already using 5.1.0 provisioner. however issue persists.