failed to rollout restart or pod stop && restart

[wwm5241]

Currently, when performing a rollout restart or stopping and restarting a pod, I frequently encounter the following error:

AttachVolume.Attach failed for volume "pvc-4479eb54-988f-4671-bd28-c1faadfaf0c1" rpc error: code = Internal desc = [ControllerPublishVolume] Attach Volume failed with error failed to attach 6f0df2f0-e937-47b2-a213-904991b0b942 volume to bcd20893-e429-4e97-a17f-8e7a947dc32 compute: Bad request with: [POST http:172.16.10.5:18774/v2.1/servers/bcd20893-e429-4e97-a17f-8e7a947dcd32/os-volume: volume 6f0df2f0-e937-47b2-a213-904991b0b942 is already attached to instances: bcd20893-e429-4e97-a17f-8e7a947dcd32"]}
This issue persists consistently.

The OpenStack version is Wallaby.
Kubernetes is running inside a VM.
When checking the attachment of the PV, it is correctly attached to the VM where the pod is running.
The PV and PVC status show as "Bound".
The reclaim policy is set to "Delete".
Could you suggest a possible resolution for this issue?

[kayrus]

thank you for a report. could you please clarify which component/pod raises these logs?

[wwm5241]

@kayrus
Hello kayrus. This issue has occurred in the GitLab, WAS, and Redis pods so far, and it continues to occur intermittently in most pods.