Disaster Recovery Migration -- SRM to Azure Site Recovery¶

Complete guide to migrating VMware disaster recovery from SRM, Zerto, and Veeam to Azure-native DR services including Azure Site Recovery, AVS DR, and JetStream DR.

DR mapping overview¶

1. Azure Site Recovery (ASR) -- replacing SRM¶

Azure Site Recovery is the primary DR solution for both AVS and Azure IaaS workloads. It provides continuous replication, automated failover, and recovery plans comparable to VMware SRM.

ASR capabilities¶

ASR for VMware VMs (on-prem to Azure)¶

Protect on-premises VMware VMs by replicating them to Azure:

# Create Recovery Services vault
az backup vault create \
  --name rsv-dr-eastus2 \
  --resource-group rg-dr \
  --location eastus2

# Configure ASR for VMware (requires configuration server on-prem)
# Deploy the configuration server OVA in VMware environment
# Register with the Recovery Services vault
# Enable replication per VM via Azure Portal

ASR for AVS (AVS to AVS or AVS to Azure)¶

For AVS workloads, ASR supports:

• AVS to AVS: replicate between two AVS private clouds in different regions
• AVS to Azure IaaS: replicate AVS VMs to Azure Managed Disks for DR
• Azure IaaS to Azure IaaS: replicate Azure VMs between regions

# Enable ASR for an Azure VM (region to region)
az vm replication enable \
  --name migrated-vm-01 \
  --resource-group rg-migrated-vms \
  --target-resource-group rg-dr-westus2 \
  --target-region westus2 \
  --target-availability-zone 1

Recovery plans¶

Recovery plans define the order in which VMs are failed over, along with pre/post scripts:

# Create recovery plan
az site-recovery recovery-plan create \
  --name rp-app-tier \
  --resource-group rg-dr \
  --vault-name rsv-dr-eastus2 \
  --groups '[
    {
      "groupType": "Boot",
      "replicationProtectedItems": [
        {"id": "/subscriptions/{sub}/.../protectedItems/vm-db-01"},
        {"id": "/subscriptions/{sub}/.../protectedItems/vm-db-02"}
      ]
    },
    {
      "groupType": "Boot",
      "replicationProtectedItems": [
        {"id": "/subscriptions/{sub}/.../protectedItems/vm-app-01"},
        {"id": "/subscriptions/{sub}/.../protectedItems/vm-app-02"}
      ]
    },
    {
      "groupType": "Boot",
      "replicationProtectedItems": [
        {"id": "/subscriptions/{sub}/.../protectedItems/vm-web-01"},
        {"id": "/subscriptions/{sub}/.../protectedItems/vm-web-02"}
      ]
    }
  ]'

The recovery plan above boots VMs in sequence: databases first, then application servers, then web servers.

Test failover¶

# Run test failover (non-disruptive)
az site-recovery recovery-plan test-failover \
  --name rp-app-tier \
  --resource-group rg-dr \
  --vault-name rsv-dr-eastus2 \
  --failover-direction PrimaryToRecovery \
  --network-id /subscriptions/{sub}/resourceGroups/rg-dr/providers/Microsoft.Network/virtualNetworks/vnet-dr-test

Always use an isolated test VNet for test failover to avoid IP conflicts with production. ASR creates test VMs in the isolated network, and you can validate application health without affecting production.

2. Zerto for AVS and Azure¶

Zerto provides near-synchronous replication with journal-based recovery, enabling any-point-in-time recovery (granularity measured in seconds).

Zerto for AVS¶

Zerto deployment on AVS¶

1. Deploy Zerto Virtual Manager (ZVM) on AVS
2. Deploy Zerto Virtual Replication Appliances (VRAs) on each ESXi host
3. Configure Virtual Protection Groups (VPGs) for VM grouping
4. Set journal history (hours/days of recovery points)
5. Configure failover networking

When to choose Zerto over ASR¶

3. JetStream DR for AVS¶

JetStream DR is optimized specifically for AVS disaster recovery scenarios.

JetStream capabilities¶

JetStream deployment¶

1. Deploy JetStream DR from AVS Azure Marketplace add-on
2. Configure protected site (source AVS) and recovery site (target AVS)
3. Define protection domains (groups of VMs)
4. Set replication policies (RPO, retention)
5. Create and test recovery plans

4. AVS stretched cluster (zero RPO)¶

For workloads requiring zero data loss (RPO = 0), AVS stretched clusters provide synchronous replication across Azure Availability Zones.

Stretched cluster architecture¶

flowchart LR
    subgraph AZ1["Availability Zone 1"]
        H1[AVS Host 1]
        H2[AVS Host 2]
        H3[AVS Host 3]
    end
    subgraph AZ2["Availability Zone 2"]
        H4[AVS Host 4]
        H5[AVS Host 5]
        H6[AVS Host 6]
    end
    subgraph Witness["Witness Zone"]
        W[vSAN Witness]
    end

    H1 <-->|Synchronous Replication| H4
    H2 <-->|Synchronous Replication| H5
    H3 <-->|Synchronous Replication| H6
    AZ1 <-->|Heartbeat| W
    AZ2 <-->|Heartbeat| W

Stretched cluster characteristics¶

5. Cross-region DR patterns¶

Pattern 1: AVS to AVS (same VMware tools)¶

Primary: AVS Private Cloud (East US 2)
    ↓ ASR or Zerto replication
DR:      AVS Private Cloud (West US 2)

Pattern 2: AVS to Azure IaaS (cost-optimized DR)¶

Primary: AVS Private Cloud (East US 2)
    ↓ ASR replication
DR:      Azure IaaS VMs (West US 2) -- only running during failover

This pattern reduces DR costs because Azure IaaS VMs are only provisioned during failover. You pay only for the replicated storage and ASR licensing, not for running VMs.

Pattern 3: Azure IaaS to Azure IaaS¶

Primary: Azure VMs (East US 2)
    ↓ ASR replication
DR:      Azure VMs (West US 2) -- auto-provisioned on failover

DR cost comparison¶

6. DR migration strategy¶

Phase 1: Assess current DR¶

Document your current VMware DR configuration:

• SRM recovery plans (count, VMs per plan, sequence)
• vSphere Replication policies (RPO settings per VM)
• Network mapping rules
• Custom scripts and callouts
• DR testing frequency and results
• RTO/RPO requirements per application tier

Phase 2: Design Azure DR¶

Map each VMware DR component to its Azure equivalent:

1. SRM recovery plans become ASR recovery plans
2. vSphere Replication becomes ASR continuous replication
3. SRM network mapping becomes ASR network mapping
4. SRM scripts become Azure Automation runbooks
5. DR testing cadence maintained or improved

Phase 3: Implement and test¶

1. Enable ASR replication for migrated VMs
2. Create ASR recovery plans mirroring SRM plans
3. Configure network mapping (source VNet to target VNet)
4. Add Automation runbooks for pre/post failover tasks
5. Run test failover for each recovery plan
6. Validate RTO/RPO meets requirements
7. Decommission SRM after successful DR test

Phase 4: Ongoing DR operations¶

• Monthly: run test failover for all critical recovery plans
• Quarterly: full DR drill with documented results
• On change: update recovery plans when VMs are added/removed
• Continuously: monitor ASR replication health via Azure Monitor

Related¶

• AVS Migration Guide
• Azure IaaS Migration Guide
• Security Migration
• Feature Mapping
• Migration Playbook
• CSA-in-a-Box Disaster Recovery

Last updated: 2026-04-30 Maintainers: CSA-in-a-Box core team