Description of problem:

Customer upgraded from version 4.12.28 to 4.14.37 where all the operators show that the upgrade was successful. However, the some internal components (like openshift-logging/elasticsearch) pods are having issue with mounting the volumes (find full error in comments):

rpc error: code = Internal desc = failed to get VolumeID from volumeMigrationService for volumePath: "[<datastore>] kubevols/<volume>.vmdk

I believe that the problem probably happened during the migration of the volumes from "in-tree" provider to csi provider.
We've checked the volumes on the vsphere and they are created and seems to be fine, just the csi driver cannot find them.

We also see that on other cluster (where the issue doesn't occur after upgrade) that the "cnsvspherevolumemigrations" is created for each PV. On the affected cluster - no "cnsvspherevolumemigrations" exists for previously created PV (like ones for elasticsearch)

Version-Release number of selected component (if applicable):

- OpenShift Container Platform 4.14.37
- on Vsphere 

How reproducible:

n/a - due to the fact we don't know what is causing the issue, there is no reproducer

Additional info:

The same issue was reported already here:
https://issues.redhat.com/browse/OCPBUGS-43804

However, due to the critical situation on customer side, I am creating new bug as I am not sure if the root cause is same. The symptoms are the same.

Will provide more info in the comments.