would seem to suggest that VBR should block the vmotion, but as you can see from this screen the snapshot was called for and then a DRS was performed (which i believe was responsible for the failure)
did VBR just not have enough time to block it (the DRS event was 1 minute 20 seconds after VBR called for snapshot)
Normally, vMotion lock is performed by the hypervisor itself after Veeam B&R invokes "VixDiskLib_PrepareForAccess" method which is described on this page of VDDK programming guide. You may ask our support engineers to examine Veeam and VMware logs to understand what happened at vSphere level after the mentioned function had been called. Please share a support case ID with us.
When Veeam has no vCenter permission rights for global access to the methods, then we can not set this flag. We continue them with the backup ignoring this. but it can lead to this situation then.
In VMWare Cloud on AWS/Outpost/Dell and Azure VMware Service this permission is not given to the customers and it is expected behaviors by design.
Support can identify the permission Limitation in the logs or identify other potential root causes.