Understanding Microsoft Azure Site Recovery for Hybrid and Private Cloud Disaster Recovery

Recent observations describe cloud development trends as follows: private cloud is behind, public cloud is ahead, and hybrid cloud is right at our feet. This article focuses on hybrid cloud disaster recovery, noting that Microsoft Hyper‑V and VMware dominate the private market and both provide private and hybrid cloud DR solutions. The discussion here centers on Microsoft’s approach, with a future article planned for VMware’s Site Recovery Manager.

Fundamental DR concepts such as Recovery Point Objective (RPO), Recovery Time Objective (RTO), and DR tiers are assumed known; readers unfamiliar can consult external resources.

Microsoft initially released a tool named Hyper‑V Recovery Manager, later renamed Azure Site Recovery (ASR) to align with Azure services. ASR supports DR between private clouds and between private clouds and Azure public cloud, with data synchronization achieved via Hyper‑V Replica or storage remote replication, and can be managed alongside System Center Virtual Machine Manager (SCVMM).

Azure Site Recovery provides near‑continuous VM replication and failover capabilities, protecting both virtual machines and data. Using ASR requires an Azure public‑cloud account to configure replication tasks and create protection, recovery, and test plans. ASR is broadly compatible, supporting Windows and Linux environments, Hyper‑V, VMware, and other physical platforms, thereby delivering hybrid cloud business continuity solutions.

Private Cloud Disaster Recovery

In private‑cloud settings, both primary and secondary sites often use storage from the same vendor, sometimes even identical models. To conserve host resources and avoid business impact, it is recommended to leverage storage‑array snapshots and replication to synchronize data between private clouds, optionally using ASR and SCVMM to establish end‑to‑end storage‑array replication, disaster, and recovery functions.

Replacing Hyper‑V Replica with SAN remote replication removes application‑level dependencies and enables holistic data protection. Storage vendors’ SCVMM SMIS Provider facilitates standard integration of arrays with SCVMM, while the ASR Provider enables VM backup to Azure or DR between two private‑cloud sites.

SMI‑S Providers are installed on a server at each site, allowing SCVMM to discover storage and perform all storage operations—such as allocating resource pools, creating Hyper‑V clusters, LUNs, replication groups, and private clouds—through the provider.

First, configure networking, create a Hyper‑V cluster, and set up replication groups and private clouds at both primary and secondary sites. Then, using the Azure account, log into the public cloud to create a recovery vault, which is required to define protection, test, and recovery plans.

If remote replication via a storage array is used for primary‑secondary data sync, install the ASR Provider at both sites to cooperate with Azure Site Recovery. Through the Azure portal, discover the storage arrays, create and map them, allow ASR to recognize the previously defined replication groups, and finally create protection tasks. ASR supports test failover, planned and unplanned failover, and reverse replication.

Hybrid Cloud Disaster Recovery

Microsoft’s hybrid‑cloud DR solution enables customers to replicate private‑cloud data to Azure public cloud, reducing investment in dedicated DR sites. The underlying storage or infrastructure is invisible to both the customer and the cloud provider, meaning data synchronization between private and public clouds cannot rely on storage‑level replication and must use Hyper‑V Replica.

Hybrid cloud is becoming mainstream, but decisions about which workloads to move to the public cloud must consider data security, the impact of data replication on business performance, and SLA requirements such as RTO and RPO.

ASR supports workflow orchestration and customization, allowing execution of custom Windows PowerShell scripts, Azure Automation runbooks, and pause points for manual intervention during failover or recovery. Recent enhancements include:

Integration of new ASR recovery plans with Azure Automation for one‑click disaster‑recovery processes.

Ability to track initial replication progress when copying VM data to cross‑region redundant Azure storage, also usable for internal cross‑site private‑cloud DR.

Simplified registration steps, removing complex certificate and integrity‑key requirements for registering on‑premises SCVMM servers with a Site Recovery vault.

Reduced replication intervals achieving near‑synchronous replication with as low as 30‑second RPO and support for consistent snapshots.

Disaster recovery is a long‑term engineering effort. While it satisfies regulatory demands, most enterprises have genuine DR needs. Beyond technology, effective DR requires proper planning, management, and decision‑making by technical leaders. Readers are encouraged to follow the ICT Architect technical exchange public account for a forthcoming comprehensive DR series.

Warm Reminder: Please search for “ICT_Architect” or scan the QR code below to follow the public account and receive more content.