We have an ESXi 6.7 host with a few guest VMs, and one of those VMs is an RDS with a single, thick provisioned 360 GB OS volume. We received an alert for free disk space on that Windows volume, and a tech attempted to expand the disk via the vmware web client larger than there was room on the datastore. (~78 GB free on the datastore, and the tech attempted to increase the disk size by 90 GB, to 450 GB total).
This action failed and produced an error in vSphere, and no additional disk space is visible from the guest VM. The issue, however, is that 70 of that 90 GB appears to have been added from the host perspective and not removed once the failure occurred, leaving us with 8 GB free on the datastore.
So from the host perspective, on the VM summary page it says that the VM has 450 GB used. The VMDK on the datastore is showing as 430 GB, however, and the HDD1 size under the VM properties says 360 GB.
The current plan is to migrate that VM to a thin provisioned disk on a local SAN, and then migrate it back to the appropriately sized thick provisioned disk and reclaim the lost datastore space on the original host.
If anyone has ever found themselves in this situation before and/or has any other ideas on how to easier resolve this issue it would be greatly appreciated.
Thanks!
