Question

Difficulty: MediumDesign Storage and Non-Relational Data HA and DR

ApexVault Pay is designing a disaster recovery strategy for its transaction ledger application. The ledger data is currently stored in a general-purpose v2 storage account configured with Read-Access Geo-Redundant Storage (RA-GRS), where West US 3 is the primary region and East US is the secondary region. The business continuity requirements specify:
- A Recovery Point Objective (RPO) of 15 minutes.
- A Recovery Time Objective (RTO) of 30 minutes.
- Read availability in the secondary region during a primary region outage must be immediate.
- Write availability must be restored in the secondary region within the RTO.
- Solution costs must be minimized by avoiding multi-region active-write configurations.

During a simulated regional outage of West US 3, the operations team needs to maintain business continuity. The application instance in East US must immediately be able to read existing ledger data. Additionally, the team plans to minimize data loss by performing a customer-managed failover to make East US the primary region.

Which of the following describes the write availability of the storage account during this recovery process?

  1. Write operations are unavailable in the secondary region until the customer-managed failover is completed, at which point the storage account is converted to locally redundant storage (LRS) in the new primary region and write capabilities are enabled.Answer
  2. B
    Write operations can be sent directly to the secondary region endpoint while the primary region is offline, before initiating the customer-managed failover.
  3. C
    Write operations are automatically replicated bi-directionally, allowing the application to write to both the primary and secondary endpoints simultaneously during the outage.
  4. D
    Write operations resume immediately in East US once the customer-managed failover is completed, while the storage account maintains geo-redundant storage (GRS) replication back to West US 3 automatically.

Answer

Write operations are unavailable in the secondary region until the customer-managed failover is completed, at which point the storage account is converted to locally redundant storage (LRS) in the new primary region and write capabilities are enabled.
The correct answer is correct because under RA-GRS, the secondary endpoint is read-only. To allow write operations in the secondary region during an outage, a customer-managed failover must be triggered. Once the failover completes, the secondary region becomes the new primary, the storage account replication tier is converted to Locally Redundant Storage (LRS), and write operations become available.

Step-by-Step Solution

1
Analyze the capabilities of the Read-Access Geo-Redundant Storage (RA-GRS) secondary endpoint during a primary region outage.
Identify that the secondary endpoint is strictly read-only prior to initiating a failover, meaning writes cannot be sent directly to the secondary region.
This establishes that write access is unavailable in the secondary region during the initial phase of the outage.
2
Determine the impact of initiating and completing a customer-managed failover on the storage account configuration.
Recall that a customer-managed failover updates the DNS records so that the secondary region becomes the new primary, enabling write access, and changes the replication type to locally redundant storage (LRS).
This identifies the state of the storage account and the replication configuration after the failover completes.
3
Compare the results against the options to find the one that accurately describes both the pre-failover write restriction and post-failover LRS transition.
The option describing write unavailability prior to failover and the subsequent LRS conversion matches these characteristics.
This confirms the correct choice based on Azure Storage DR design rules.

Key Concept

Azure Storage customer-managed failover behavior and replication state changes for GRS/RA-GRS accounts.
Rate this question