Azure Site Recovery tutorial for workload disaster recovery

What is Azure Site Recovery?

As the name indicates, Azure Site Recovery is a service based in Microsoft's Azure cloud platform. Customers can replicate their on-premises Windows and Linux VMs running on VMware or Hyper-V and physical Windows or Linux servers to Azure. In the event of a disaster, such as a power outage or hardware failure, VMs fail over and continue to operate in the Azure cloud to minimize downtime.

An agent installed on the physical server or virtual machine facilitates the replication process in Azure and assists with tasks related to configuration, monitoring and troubleshooting.

Azure Site Recovery supports application-aware replication for Microsoft products, such as Exchange, SQL Server, Active Directory and SharePoint, and major vendors, including Oracle, SAP and IBM.

Azure Site Recovery also replicates Azure VM workloads.

Organizations can tailor Azure Site Recovery for their specific needs and customize the recovery time objective (RTO) and recovery point objective (RPO) for more frequent replication of important workloads to avoid data loss. Advanced settings also offer a way to build a recovery plan for an orchestrated failover and use of automation via PowerShell scripts to accelerate DR efforts.

What are the benefits of Azure Site Recovery?

One of the real advantages of this Azure service for a Microsoft-based shop is integration. All the functionality is built into the admin portal and requires little effort to configure beyond the agent installation, which can be done automatically. Offerings from other vendors, such as Zerto and Veeam, work the same way but require additional configuration using a management suite based outside the Azure portal.

As a cloud-based service, Azure Site Recovery is under constant development. Microsoft introduced Azure Site Recovery in 2014 and issues rollups nearly every month to add features or improve functionality with the service and related software.

What is Azure Site Recovery pricing?

One of the big issues for any platform is cost. Microsoft charges based on the number of protected instances. Organizations interested in testing the service can use it for free for the first 31 days.

Each protected instance to Azure costs $25 per month with additional fees for the Azure Site Recovery license, storage in Azure, storage transactions and outbound data transfer. For a protected instance to a customer-owned site, the cost is $16 per instance.

As with most systems, there are caveats, including how replication and recovery are tied to specific Azure regions depending on the location of the cluster.

Microsoft does not charge for incoming replication, but does charge for outgoing replication. The total cost depends on the Azure region and how much egress data is transferred. Each attached disk to the instance adds to the charges with faster disk types costing more.

During a failover, Microsoft adds a compute charge for the VMs that are running in Azure.

What are Azure Site Recovery prerequisites and requirements?

Using Azure Site Recovery requires an active subscription to Azure.

In Azure, there must be an Azure virtual network to place the VMs in a failover and a storage account to hold replicated data.

Replication for VMware VMs and physical servers requires a configuration server to handle replication to Azure. The server is set up as a highly available VMware VM and uses a registration key from the recovery vault for access.

How does Azure Site Recovery work with cloud workloads?

Microsoft integrates Azure Site Recovery with cloud workloads to make protection of those instances easy. To add Azure Site Recovery to a single VM, click the DR option on the VM pane, select the region to recover to and accept the defaults.

While this is a quick way to set up DR for the single VM, it isn't best practice because it ignores a lot of the advanced options and configuration settings. Admins who might have been overwhelmed by earlier versions of Azure Site Recovery that were overly complicated and difficult to deploy will appreciate recent updates that simplify the process.

After setting up Azure Site Recovery to protect a VM, the Azure subscription objects section now has a Recovery Services vault that ends in -asr. Microsoft creates this vault when you use the simpler method.

How to work with the recovery vault

The ability to fail over in a consistent manner is a significant benefit that gives the enterprise more fine-grained control over groups of VMs.

Ideally, the admin sets up a recovery vault -- or several -- in advance to group applications for a more manageable and consistent failover experience.

To create a recovery vault, click Create a resource from the Azure portal, search for Recovery Services vault and then select Create to build the vault, following the configuration prompts to give it a name, resource group and Azure region.

To add VMs to a recovery vault, use the Disaster Recovery button on the Recovery Services vault and blade, and select the new vault.

The vault provides options to manage the backup polices for stored VMs, provides the restore function from a recovery point in the VM and monitors the replication status of protected VMs.

What are some potential issues with Azure Site Recovery?

There are some potential issues during the Azure Site Recovery setup process, but they are usually easily diagnosed.

If the VM has just been provisioned, it may still be completing tasks in the background, such as installing agents, so try again after several minutes if it doesn't appear in Azure Site Recovery.

If the VM is on premises, then make sure the Azure Site Recovery agent is installed. It may be prudent to reboot and try to add it to Azure Site Recovery again.

Communication issues are by far one of the most common issues. Microsoft provides lengthy documentation on setup requirements and implementation for on-premises VMs to Azure. It does take some time to make sure everything is properly configured.

Stuart Burns is a virtualization expert at a Fortune 500 company. He specializes in VMware and system integration with additional expertise in disaster recovery and systems management. Burns received vExpert status in 2015.