I've tried searching through the community but couldn't find a similar thread so forgive me if this has been answered before 
I'm currently running some tests on the vSphere Replication Appliance.
My setup is as follows(all running 60 day trial licenses):
ESXi1.domain.local - ESXi 5.5
ESXi2.domain.local - ESXi 5.5
DSS1.domain.local - Open-e DSS Lite (Primary shared storage)
DSS2.domain.local - Open-e DSS Lite (Shared storage for replication)
vc.domain.local - vCenter 5.5 (running from DSS2)
vra.domain.local - vSphere Replication Appliance 5.5 (running from DSS2)
Nostalgia VM - Test VM (running from DSS1)
Both ESXi hosts are in same Datacenter (ROE_CENTER) and Cluster, and both hosts has read/write access to the two shared storages.
I have no problems configuring replication of Nostalgia (VRA_Before.png):
Destination storage: DSS2
RPO: 4 hours
POI: Disabled
When I simulate a total crash of the primary storage (DSS1.domain.local) by pulling the power, the Nostalgia VM is still seen as powered on by the ESXi hosts for some reason, so I manually power it off (right click VM -> Shut Down Guest OS)
Opening the vSphere Replication plug-in in the vSphere Web Client and going to the Monitor tab, the name of the Nostalgia VM is now displayed as the path to the VM on the datastore ( %2fvmfs%2fvolumes%2f<UID>%2fNostalgia%2fNostalgia.) instead of the original VM name.
I can right-click the VM, recover it, to a Recover-folder, using "Recover with latest available data" and then find, and power on, the VM in the vCenter inventory.
But still with the datastore-path name (VRA_After.png)
Only the display name is incorrect. Physical filelocations on the replication store (DSS2) are still in the correct Nostalgia directory.
Isn't it possible to recover a VM, keeping the correct name, when source datastore is offline, using the vSphere Replication Appliance within the same vCenter?
Or should I instead look into recovering replicated VMs by renaming VM-files on the datastore and registering them correctly using PowerCLI scripts?