The answer is D. its just a little confusing the way the question is written: https://helpcenter.veeam.com/archive/backup/95/vsphere/replica_seeding_vm.html Replica seeding can be used if you have a backup for the replicated VM on the backup repository located in the DR site. In this case, you can point the replication copy job to this backup. During the first session of the replication job, Veeam Backup & Replication will use this backup file as a "seed". Veeam Backup & Replication will restore the VM image from the backup and register the VM replica on the target host. After that, Veeam Backup & Replication will synchronize the VM replica with the source VM. All subsequent incremental replication runs will be performed in the regular manner.

D is incorrect because Replica Seeding reduce size of data for only first run https://helpcenter.veeam.com/archive/backup/95/vsphere/replica_name_vm.html As per Veeam documentation, if we understand how WAN Accelerator works https://helpcenter.veeam.com/docs/backup/vsphere/wan_hiw.html?ver=120 B is the correct answer.

Replica mapping If a replica for the VM that you plan to replicate already exists on the target host in the DR site, you can use replica mapping. Replica mapping helps reduce the amount of VM data transferred over the network. To use replica mapping, you must point the replication job to a VM replica on the host in the DR site. During the first session of the replication job, Veeam Backup & Replication will calculate the difference between the source VM and VM replica and copy necessary data blocks to synchronize the VM replica to the latest state of the source VM. All subsequent incremental replication sessions will be performed in the regular manner.”

B makes sense to me for subsequent runs of the job.

It could also be B. The book is not clear.

Shouldn't this question say "over the first run"? That's the whole idea of seeding.