Hyper-V Replica Pros and Cons

Each technology has its pros and cons and Hyper-V Replica is no exception. In this article we compare BackupChain virtual machine replication with Hyper-V Replica. Both technologies have some overlap and common use scenarios but differ in how they work.

Hyper-V Replica Pros and Cons

Hyper-V Replica is integrated in Hyper-V. It can be quite difficult to set up. It can be quite a challenge to monitor it properly. We have seen it countless times. Customers think they have a ‘backup’ (note: Hyper-V Replica is absolutely not a backup) and then they realize their Hyper-V Replica wasn’t working for months and no one noticed!
So in terms of reliability, there is some babysitting involved with Hyper-V Replica, or at least the IT admin has to install a monitoring system to make sure it doesn’t just silently quit working.
Hyper-V Replica works at the block-level. It’s file system, service, and application agnostic. The blocks are simply moved from A to B. That’s another reason why Hyper-V Replica is not a backup. In a live system, you can’t just take blocks and copy them because the services and applications keep data in RAM.
You can’t also take a Hyper-V Replica, just like that, and work with it. You would have to make sure the replication is up-to-date and finalized. Naturally it’s good to have a replica when the source VM is no longer around, but having a proper, application consistent backup is much more desirable because it’s a real and proper copy of the VM, not a ‘puzzle’ of blocks that was stitched together.

However, Hyper-V Replica works well for very large VMs because of its design. When you have a multi-terabyte VM, it can be efficiently replicated over a very slow link, and quite often, if the number of changed blocks is minimal.
Hyper-V Replica only works with a narrowly defined target. You can’t send your replicas wherever you want, for example to a Linux box or a simple NAS share. There has to be a Hyper-V at the receiving side, and the authentication requirements can be quite complex to set up and manage when you are dealing with remote sites.

BackupChain Replication Pros and Cons

BackupChain replication is very simple to set up. You can use any storage as a target: another disk, another server, a Linux share, a NAS, SAN, and even FTP sites. It’s a very straightforward setup.

When a live backup takes place, as with BackupChain, the services and the OS are instructed to prepare for live backup, which means they have time to write everything to disk properly. After the disk data structures are kept in a consistent state, the backup copy initiates. This also involves a lot of work because the applications have to prepare their disk structures. This does not occur with Hyper-V Replica; ergo, when more work is done it will also take a little more time.

With BackupChain you know that a full copy is made properly. The old copy is kept until the new copy is finished successfully. This guarantees that you always have a good copy elsewhere. Plus multiple copies can be kept at the target if you want. Such guarantees cannot be made with Hyper-V Replica.
In addition, BackupChain logs everything and checks every step along the way, and supports complex scheduling. You can always track what happened, and if something fails you can quickly find the cause. We have seen it countless times that Hyper-V replication stops working and no one notices; this can’t happen with BC because you can also set up various alerts in case of errors.

Another issue with Hyper-V replication is that it can result in a corrupt copy if it’s not working properly, fast and often enough. Imagine a very large VM is replicated over a very slow link. It will never work properly if too much data is changed too fast. And will you be made aware of it? How?

With BC, on the other hand, you know the process takes a certain time to finish, and when it’s done, it’s 100% complete. Even if the VM is very large, the copy that BC makes will finish eventually and will be consistent. So if the situation occurs that too much data changes too fast, this situation will have no effect on how the BackupChain replication works because you will get a full consistent copy each time. For large VMs this process may take a while over slow links, and this is where the two technologies differ in timing; however, VM size is just relative to the transport medium. When you switch to a much faster network and storage system, a large VM suddenly does not appear to be as large anymore.
Hyper-V replication can work well, too, but requires active monitoring and has its own ‘side-effects’ one needs to be aware of, such as that the replica is actively being written to all the time or within short intervals. There is no clear-cut, predictable synchronization taking place, as with BackupChain.
Also when BackupChain takes the copy, it’s an application consistent copy, guaranteed to be consistent when you boot it. BackupChain works together with Exchange, Windows, NTFS, SQL Server etc, to ensure a good copy is made. Hyper-V Replica doesn’t do that as it is a continuous, block-level, file system, operating system, and service-agnostic process.

Download BackupChain    Learn More About BackupChain