You can perform three types of failovers with Hyper-V Replica after it is configured: planned failovers, unplanned failovers, and test failovers. It’s somewhat likely you’ll see an exam question in which you need to understand the difference among them and when they are used.
A planned failover is the only failover you initiate from the primary server. You use this method whenever you can manually shut down the primary VM, and the primary and replica servers can still communicate.
A planned failover is the preferred failover type because no data is lost. In fact, you cannot even use this option to fail over to the latest recovery point or to any earlier recovery point. With a planned failover, only an exact copy of the current primary VM and its VHDs can be failed over to the replica server.
A planned failover is a good option in the following situations:
– You want to perform host maintenance on the primary server and temporarily want to run the VM from the replica.
– Your primary site is anticipating a possible power outage and you want to move the VM to the replica site.
– You are expecting a weather emergency, such as a flood, and you want to ensure business continuity.
– Your compliance requirements mandate that you regularly run your workloads for certain periods of time from the replica site.
To perform a planned failover, you begin by shutting down the primary VM. You then right-click the VM in Hyper-V Manager, click Replication, and then click Planned Failover, as shown in Figure 3-39. The latest updates are then sent to the replica server, the VM is failed over, and the replica VM is automatically started on the remote server. At the end of this operation, the replication relationship is reversed, so what was the replica server becomes the primary server, and vice versa.
FIGURE 3-39 Performing a planned failover from the primary server.
This type of failover is called an unplanned failover in the Windows Server 2012 and Windows Server 2012 R2 documentation, but in the actual interface, it’s called just “failover.” On the 70-412 exam, you might see it referred to either way.
An unplanned failover is performed at the replica server. You perform this failover type when the primary VM fails suddenly and cannot be brought back online. An unplanned failover is a good option in the following situations:
-Your primary site experiences an unexpected power outage or a natural disaster.
-Your primary site or VM has had a virus attack, and you want to restore your business quickly with minimal data loss by restoring your replica VM to the most recent recovery point before the attack.
To perform an unplanned failover, in Hyper-V Manager on the replica server, right-click the replica VM, click Replication, and then click Failover, as shown in Figure 3-40.
FIGURE 3-40 Performing an unplanned failover on the replica server
When you perform an unplanned failover, you have to choose a recovery point, as shown earlier in Figure 3-34. The VM is then started on the replica server.
After the replica VM is started, the replica relationship with the primary VM is broken, and replication stops. If at some later point you can bring the original primary VM online, you can resume replication by reversing the replication relationship. After you perform this operation,the local replica server becomes the new primary, and the remote primary becomes the new replica. To reverse replication in this way, right-click the VM on the replica server, click Replication, and then click Reverse Replication, as shown in Figure 3-41. This step starts the Reverse Replication Wizard, which allows you to reenter the settings for the replica.
FIGURE 3-41 Reversing replication
Another option you can see on the Replication submenu in Figure 3-41 is Cancel Failover.You can safely choose this option after you perform an unplanned failover as long as no changes have been made to the replica. After you cancel a failover, you have to manually resume replication on the primary VM by right-clicking it and selecting Resume Replication.Cancelling a failover is a good idea if you quickly discover after performing an unplanned failover that the primary VM can be brought online.
Important: Remember the Reverse Replication option and the Cancel Replication option for the exam.Remember the Reverse Replication option and the Cancel Replication option for the exam.
A test failover is the only failover operation you can perform while the primary VM is still running. The purpose of this failover type is to simulate an unplanned failover so that you can ensure that it will function as planned in case of an emergency.
To perform a test failover, in Hyper-V Manager on the replica server, right-click the replica VM, click Replication, and then click Test Failover. You then have to select a recovery point, just as you do with an unplanned failover. Next, a local, disposable copy of the replica VM is created on the replica server. The new copy of the VM appears in Hyper-V Manager in a stopped state with the tag “- Test.” For example, a test failover of a VM named “MyVM1”would result in a new VM called “MyVM1 – Test”. You can then start the new VM manually to see if it works as expected.
By default, the virtual network adapters of the test VM are disconnected from all virtual switches. If desired, you can preattach the adapter(s) of the test VM to a virtual switch of your choice. To do so, open the settings of the base replica VM, expand Network Adapter, and then click Test Failover, as shown in Figure 3-42. Make sure you choose a virtual switch that will not create any conflicts in a production network.
After you examine the functioning of the test VM, you can safely delete it in Hyper-V Manager.
FIGURE 3-42 Preattaching the network adapter of a failover test VM to a virtual switch