Rollover Update for a deployment that uses ReadWriteOnce got failure due to the second deployment pod getting started on the second node.
I have followed dynamic provisioning of persistent volume using the doc: https://kubernetes.io/docs/concepts/storage/dynamic-provisioning/
The pods are created during the first deployment but when the image is replaced, the second deployment shows the following error as the pod is being created in the next node.
Multi-Attach error for volume "xxxxxxxxxxxxxxxxxxx" Volume is already used by pod(s) "xxxxxxxxxxxx"
I have tried the strategy of using recreate but there would be downtime as the new pod will be created only after termination of previous pod. Reference: Updating a deployment that uses a ReadWriteOnce volume will fail on mount
I have also tried using ReadWriteMany along with NFS persistent volume. But this approach is not suitable for our use case as the solution needs to be multi cloud solutiun
Is there any approach with access mode ReadWriteOnce that can be used along with rolling update with no downtime in gcp
