Replication error after svMotion to different datastore

[RooDBwoY]

Afternoon all,

Tried searching around for this error, with no joy and wondered if anyone else has seen something similar. It seems to present itself while on our remote Veeam server (over a reliable fibre 100mbit WAN link with excellent latency properties) if I have svMotioned the VM (in it's entirety or only certain disks) onto a different SAN datastore.

Veeam version: 5.0.2.230 (64 bit) [with the VMFS 5 update applied)
OS: Windows Server 2008 SP2
vCenter Server v4.1 at both ends (latest build)
Speed: Usually sits at 12MB/sec solidly
Replicas store: 3
Replication type: 'Pull' over dedicated 100mbit WAN link

Error: "Invalid VM configuration. Invalid operation for device '0'."

The task refuses to replicate and the only way to ever get it to successfully replicate again is to remove the entire replica and start again. This did not go down too well when I discovered it was on a 1.2TB VM!

Any ideas anyone? I don't want to have to be ultra-wary of moving machines around our live infrastructure for fear of them leaving us back at square one at our remote location.

Cheers,

James.

[tsightler]

Are you talking about moving the source VM or the target VM?

[RooDBwoY]

Sorry, I wasn't clear. We have our entire VMware infrastructure locally (with a number of SAN datastores) and a vCenter / Veeam server at our remote location, which is performing the pull replication. If I move a VM on our local infrastructure to a different SAN datastore the replication delta jobs fail with the above error the next time the job runs. The only way to rectify it is to scrap off the entire replica and start again from scratch.