Question

Difficulty: HardDesign Azure Site Recovery Solutions

An organization is designing a disaster recovery (DR) solution for a smart energy analytics platform hosted on Azure virtual machines in the UK South region. The target DR region is UK West. The platform consists of three tiers:

* Web Tier: Two virtual machines. Each virtual machine has one Standard SSD operating system (OS) disk with a write churn of 0.5 MB/s0.5\text{ MB/s}.
* Application Tier: Two virtual machines. Each virtual machine has one Standard SSD OS disk with a write churn of 1.5 MB/s1.5\text{ MB/s}.
* Database Tier: Two virtual machines running SQL Server Enterprise. Each virtual machine has one Standard SSD OS disk (2 MB/s2\text{ MB/s} churn), two Premium SSD (P30) data disks (24 MB/s24\text{ MB/s} churn per disk), and one Premium SSD (P20) log disk (28 MB/s28\text{ MB/s} churn).

The solution must meet the following business requirements:

* The web and application tiers must have a Recovery Point Objective (RPO) of 2 hours2\text{ hours} and a Recovery Time Objective (RTO) of 4 hours4\text{ hours}.
* The database tier must have an RPO of 10 seconds10\text{ seconds} and an RTO of 15 minutes15\text{ minutes}.

Which two recommendations should you include in the disaster recovery design? (Select two.)

  1. Replicate the web and application tier virtual machines to the target region by using Azure Site Recovery.Answer
  2. Configure SQL Server Always On Availability Groups to replicate the database tier to the target region.Answer
  3. C
    Replicate the database tier virtual machines to the target region by using Azure Site Recovery with a high-churn replication policy.
  4. D
    Configure Azure Backup with Cross-Region Restore (CRR) to replicate the database tier virtual machines.

Answer

Replicate the web and application tier virtual machines to the target region by using Azure Site Recovery, and configure SQL Server Always On Availability Groups to replicate the database tier to the target region.
The correct disaster recovery design uses Azure Site Recovery for the web and application tier virtual machines because their write churn is low and the 2 hours2\text{ hours} RPO and 4 hours4\text{ hours} RTO are fully supported. For the database tier, native SQL Server Always On Availability Groups with asynchronous replication must be configured. This meets the 10 seconds10\text{ seconds} RPO and bypasses Azure Site Recovery limits, as the database disks exceed standard ASR churn thresholds of 20 MB/s20\text{ MB/s} per disk and 54 MB/s54\text{ MB/s} per VM.

Step-by-Step Solution

1
Analyze the RPO and RTO requirements for both the web/application tiers and the database tier.
The web/application tiers require an RPO of 2 hours2\text{ hours} and RTO of 4 hours4\text{ hours}. The database tier requires an RPO of 10 seconds10\text{ seconds} and RTO of 15 minutes15\text{ minutes}.
Identifying the recovery objectives determines which replication technologies can meet the constraints.
2
Evaluate the capability of Azure Site Recovery (ASR) to meet the RPO of the database tier.
ASR replication frequency for Azure VMs generates recovery points every few minutes, failing to support a 10 seconds10\text{ seconds} RPO.
This rules out ASR for the database tier and mandates a native replication mechanism.
3
Verify write churn rates of the database tier disks against ASR limits.
The database disks (24 MB/s24\text{ MB/s} and 28 MB/s28\text{ MB/s}) exceed standard ASR disk-level thresholds (20 MB/s20\text{ MB/s} for P20/P30), and the total VM write churn (78 MB/s78\text{ MB/s}) exceeds the standard 54 MB/s54\text{ MB/s} limit.
Verifying limits confirms that ASR is technically unsuitable for this database workload without special policy modifications, and even then fails the RPO constraint.
4
Evaluate the replication options for the web and application tiers.
The web and application VMs have very low write churn and their RPO/RTO requirements are fully satisfied by ASR.
Using ASR for stateless application layers minimizes administrative and replication configuration overhead.

Key Concept

Disaster recovery design involving write churn limits and RPO/RTO constraints in Azure Site Recovery.
Rate this question