In the article OMS Azure Site Recovery: solution overview Azure Site Recovery characteristics were presented and examined aspects that make it an effective and flexible solution for creating business continuity and disaster recovery strategies for your data center. In this article we will see how to use Azure Site Recovery to migrate even potentially heterogeneous environments to Microsoft Azure. Increasingly we are opposite the ’ need not only to create new virtual machines in Microsoft public cloud, but also to migrate existing systems. To perform these migrations you can adopt different strategies, including also appears Azure Site Recovery (ASR) that allows us to easily migrate virtual machines on Hyper-V, VMware, physical systems and workloads of Amazon Web Services (AWS) to Microsoft Azure.
The following table shows what is possible migration scenarios deal with ASR:
| Source | Destination | Supported Guest OS type |
| Hyper-V 2012 R2 | Microsoft Azure | All supported guest OSS in Azure |
| Hyper-V 2008 R2 SP1 and 2012 | Microsoft Azure | Windows and Linux * |
| VMware and physical servers | Microsoft Azure | Windows and Linux * |
| Amazon Web Services (Windows AMIs) | Microsoft Azure | Windows Server 2008 R2 SP1 + |
* Limited support to Windows Server 2008 R2 SP1 +, CentOS 6.4, 6.5, 6.6, Oracle Enterprise Linux 6.4, 6.5, SUSE Linux Enterprise Server 11 SP3
When you need to perform migration task is usually critically important respect the following points:
- Minimize downtime of production workloads during the migration process.
- Have the opportunity to test and validate the solution works in the target environment (Azure in the specific case) before the migration.
- Make a single migration of data useful to the validation process for the actual migration.
With ASR this is possible by following this simple flow of operations:
Figure 1 – Migration flow with ASR
Let us now see in detail what are the operations to be carried out in a migration scenario of virtual machines on a Hyper-V host 2012 R2 to Microsoft Azure.
First, since you have to create an Azure Portal Recovery Service Vault in the subscription to which you want to migrate virtual machines:
Figure 2 – Creating Recovery Service Vault
Afterwards you must prepare l ’ order to use Azure infrastructure Site Recovery. All you can do so by following the wizard proposal from the Azure Portal:
Figure 3 – Infrastructure preparation
After declaring your migration scenario (virtual machines on Hyper-V is not managed by SCVMM to Azure), assign a name to the site Hyper-V and agree to it l ’ Hyper-V host that holds the virtual machines:
Figure 4 – Preparing Source: step 1.1
Figure 5 – Preparing Source: step 1.2
At this point you need to install on ’ Hyper-V hosts the Microsoft Azure Site Recovery. During the installation ’ you can specify a proxy server and your registration key to the vault, which you need to download it directly from the Azure Portal:
Figure 6 – Provider installation ASR
Figure 7 – Configuring access to the vault
Figure 8 – Proxy settings
Figure 9 – Registration vault ASR
After waiting a few moments on the Hyper-V server registered at Azure vault Site Recovery will appear on the Azure Portal:
Figure 10 – Preparing Source: step 2
The next step requires you to specify on which subscription Azure will create virtual machines and the deployment model (Azure Resource Manager – ARM in the following case). At this point it is important to verify that a storage account and a virtual network that attest to the virtual machines:
Figure 11 – Target Preparation
The next step is where you specify which replication policy associate with the site. If there are no previously created policy you should configure a new policy by setting the following parameters, best suited to your environment:
Figure 12 – Setting replication policy
Figure 13 – Replication policy Association
L ’ last ’ ’ steps of preparation of infrastructure require the implementation of the Capacity Planner, very useful tool to estimate bandwidth usage and storage l ’. It also allows you to evaluate a series of other aspects that you need to take well into account replication scenarios to avoid problems. The tool can be downloaded directly from the Azure Portal:
Figure 14 – Capacity planning
At this point you have completed all the configuration and preparation of ’ infrastructure and you can continue selecting which machines you want to replicate from site previously configured:
Figure 15 – Enabling replication
In the next step you can select the replicated machine configurations in terms of Resource Group, Storage Account and Virtual Network – Subnet:
Figure 16 – Target recovery settings
Between all the machines that are hosted on the ’ Hyper-V you should select to whom you want to enable replication to Azure:
Figure 17 – VMs selection to be replicated
For each selected virtual machine, you must specify the guest operating system (Windows or Linux), What ’ is the disk that holds the operating system and what data disks you want to replicate:
Figure 18 – Properties of VMs in replica
After completing the configuration of all steps will begin the replication process according to the settings configured in the policy specifies:
Figure 19 – Replication steps
After the initial replication is recommended to verify that the virtual machine still works correctly in Microsoft Azure environment by placing an “Test Failover” (point 1 Dell ’ image below) and after appropriate checks should be “Planned Failover” (point 2) to have the virtual machine available and ready to be used in production environment. When this is done can be considered completed the migration to Azure of your system and you can remove the replication configuration all ’ within the Recovery Service Vault (point 3).
Figure 20 – Finalization of the migration process
Conclusions
Azure Site Recovery with simple guided steps allows us to easily migrate, safely and with minimum downtime systems that are located in our datacenter or workloads found in Amazon Web Services (AWS) to Microsoft Azure. I remind you that the functionality of Azure Site Recovery can be tested by activating a trial of environment Operations Management Suite or of Microsoft Azure.
