.shock - Fotolia

What did Microsoft add to Hyper-V Replica in 2012 R2?

Microsoft updated its Hyper-V Replica feature for Windows Server 2012 R2, but do you know how these changes can affect your RTO and recovery options?

Hyper-V Replica provides replication services for virtualized workloads running on Windows Server 2012 (and later) Hyper-V hosts. In its first release with Windows Server 2012, Hyper-V Replica provided a number of great backup and recovery features. Microsoft’s main focus has been to provide replication features that will help reduce the overall recovery time objective (RTO) for the virtualized workloads and help in workloads at both primary and replica sites.

Reduction in overall RTO: When using Hyper-V Replica, the primary objective is to reduce the time it takes to restore business services. Restoring business continuity depends on two things: How fast a replica VM can come online and how much data you would need to restore to bring replica VM up to date. Microsoft provides two new features in Hyper-V 2012 R2 that would help you reduce the overall RTO:

  • Generation 2 VMs and Hyper-V Replica: Microsoft introduced Generation 2 VMs in Hyper-V 2012 R2. Since Generation 2 VMs offer faster boot, it helps you bring up the replica VM as quickly as possible. The replica VM is turned off at the replica site, it is brought online during the replication cycle in case of any disasters with primary site or VM. It takes less time to bring a Generation 2 replica VM online than it  takes to boot up a Generation 1 VM.
  • Replication frequency: In Hyper-V running on Windows Server 2012, the replication interval was five minutes, and  was not configurable. In Hyper-V 2012 R2, you can select between 30 seconds, five minutes and 15 minutes as the replication interval. If you run an application inside the primary VM that performs write operations every few seconds, such a SQL Server, setting a replication interval of 30 seconds will ensure that changes made at the primary VM are replicated to the replica VM as quickly as possible which, in turn, helps reduce the time it takes to recover a VM. For example, if changes are replicated every 30 seconds, you wouldn’t have to spend much time in bringing the replica VM up to date. You could survive 30 seconds data loss, but recovering five minutes of data might take some time. You might not want to set replication frequency of 30 seconds for all everything, but the replication interval can be set per VM.

Recovery copies: Recovery copies are generated every hour. You can keep a maximum of 15 recovery copies in Windows Server 2012 and was increased to 24 in Hyper-V 2012 R2. This means you can recover a VM from a recovery point that was created 24 hours prior.

Extra layer of protection: Microsoft introduced the extended replication feature in Hyper-V running on Windows Server 2012 R2 hosts. With the extended replication feature, you can extend replication of a VM to a third site which allows you to recover a VM in case both the primary and replica servers are impacted by a disaster. All changes that occur at the primary server are copied to the replica server and the replica server, in turn, copies these changes to the extended replica server.

It is worth mentioning that Hyper-V 2012 R2 supports replication for newer Linux distributions and provides VHD-level consistent snapshots that you can use to recover a Replica VM running a Linux operating system.

Next Steps

How does a Hyper-V Replica failover work behind the scenes?

Answers to your questions about Hyper-V Replica

When and when not to use Hyper-V Replica

Dig Deeper on Microsoft Hyper-V and Virtual Server