im looking for assistance to Set up disaster recovery for Azure VMs

For Azure VMs, use Azure Site Recovery (ASR) for regional disaster recovery. Azure Backup should still be enabled, but it solves a different problem: Backup restores deleted or corrupted data, while Site Recovery continuously replicates the VM and orchestrates failover.

A practical setup sequence is:

1. Define the required RPO/RTO and choose a target Azure region. Verify that the VM OS, disk types/sizes, encryption configuration, and target-region VM quota are supported.
2. Prepare the recovery network: target VNet/subnets, NSGs, DNS, routing, load balancers, and—if capacity is critical—a Capacity Reservation Group.
3. Create or select a Recovery Services vault, open Site Recovery > Azure virtual machines > Enable replication, and choose the VM, target resource group, network, replica disks, and replication policy.
4. For a multi-tier application, create a recovery plan so services start in dependency order—for example database, middleware, then web tier. Include DNS, secrets, PaaS dependencies, and any manual steps ASR does not replicate.
5. When replication is healthy, run a Test failover into an isolated VNet. Validate application access and data, then clean up the test. This does not interrupt production replication.

Document planned/unplanned failover, reprotection, and failback, and repeat the drill at least quarterly and after major changes.

Microsoft guides: enable Azure-to-Azure replication and run a test failover.

Hello Danson Ndetei

When you enable replication for an Azure VM, the Site Recovery Mobility service extension is installed on the VM and registers it with ASR. Disk writes are sent to a cache storage account in the source region, replicated to the target region, and used to create recovery points. During failover, a recovery point is used to spin up the VM in the target region.

Note: Azure-to-Azure DR is not supported for Recovery Services vaults created in South Central India and North East US 5 regions.

Step 1: Verify Prerequisites & Permissions:

• Supported regions – Review supported source/target regions.
• VM OS – Confirm Windows/Linux OS and VM compute, storage, and networking requirements are supported.
• Permissions needed:
  • Create vault → Admin/Owner on subscription.
    • Manage Site Recovery → Site Recovery Contributor role.
      • Create failover VMs → Virtual Machine Contributor role (or granular rights on VNet, Storage, Managed Disk).
      • Target region capacity – Ensure quota supports matching VM sizes.

Step 2: Prepare the Source VMs:

• Ensure outbound connectivity to required Site Recovery URLs / Service Tags.
• Install the latest root certificates on Windows/Linux VMs.
• Avoid authentication proxies for outbound traffic.

Step 3: Create a Recovery Services Vault

1. In the Azure portal → search Recovery Services vaults → + Create.
2. Choose subscription, resource group, vault name, and a target region (must differ from source).
3. Click Review + Create → Create.

Step 4: Enable Replication.

Option A – From the Vault:

1. Open the vault → Site Recovery page → under Azure virtual machines, click Enable replication.
2. Source: select source region, subscription, resource group, deployment model, and choose whether zone-to-zone DR is required.
3. Virtual machines: select up to 10 VMs at a time.
4. Replication settings: target location, target RG, VNet, storage, replication policy. Recommendation: keep target location same as the vault.
5. (Optional) Create a replication group for multi-VM consistency.
6. Click Enable replication.

Option B – From the VM blade (Quickstart):

1. Go to the VM → Operations → Disaster recovery.
2. Select the Target region → Review + Start replication (or Advanced settings to customize).
3. Click Start replication.

For VMs with high data change rate, enable the High Churn option (uses Premium Block Blob storage). Premium SSD v2 / Ultra Disks use High Churn by default.

Step 5: Monitor & Test Failover

• Track replication health from the vault → Replicated items page.
• Run a Test failover (uses an isolated network — no impact to production).
• After validation, run Cleanup test failover to remove test resources.
• Failover types available:
  • Planned – graceful, minimal data loss.
    • Unplanned – uses latest recovery point for unexpected outages.
      • Test – validation only.

References:

Could you please share us the below details?

1. What is the source region and desired target region?
2. Is this for region-to-region or zone-to-zone DR?
3. How many VMs need replication, and what is their OS / disk type (Standard SSD, Premium SSD v2, Ultra)?
4. Do any VMs have high churn (>20 MB/s write rate)?
5. Are the VMs encrypted (ADE / CMK)? — different flow applies.
6. Is the target region VNet / Subnet / NSG already in place, or does it need to be created?
7. Preferred RPO/RTO and retention policy?

Thanks,

Suchitra.