A: There's no hard limit for the number of Hyper-V Replica servers that can be enabled from a Hyper-V server or to a Hyper-V server. However, you should carefully consider the following when planning the number of Hyper-V replicas and the infrastructure required.

First, for each Hyper-V replica there's additional write activity on the volume containing the virtual hard disks (VHDs) of the virtual machine (VM) for the log files. The write activity isn't doubled, as there are many optimizations used in the log creation. However, it still increases, around 20 percent is fairly typical from my tests, so this would need to be considered and would vary based on the characteristics of the workload running inside the VM.

Second, and still related to storage, is the actual space the log files consume. Normally the log files are sent to the replica every five minutes. After confirmation that the log file has been played into the replica VM, the log file is deleted from the primary. However, if there's a large amount of change on the source VM (also called churn), the log file could grow large, and any interruption in the transmission of the log file would result in multiple log files queuing up on the source server, consuming more disk space.

Third, the log files have to be sent over the network, and depending on the workload of the VM, the log files could be large and require network bandwidth to be sent. The more VMs enabled for Hyper-V Replica, the more bandwidth required. If the network bandwidth between the primary and the replica isn't sufficient, the log files will start to queue up and the replica will start to creep further out of synchronization.

Finally, and a much smaller consideration, Hyper-V Replica will use some additional CPU and memory. However, I haven't seen this to be a big consideration, and the disk and network factors would be more important.