AZ-305 Question 70
Single answerAn online retailer currently hosts its product catalog in Azure SQL Database using the General Purpose tier. As demand has grown, peak traffic periods generate significant read workloads that are slowing down response times. The team wants a scalable solution to handle these spikes in read requests with minimal application code changes. Which recommendation best addresses this requirement?
- A
Upgrade the Azure SQL Database to the Business Critical tier and use built-in read-scale replicas
- B
Implement manual sharding by splitting the database across multiple Azure SQL Database servers
- C
Migrate the data to Azure Cosmos DB with multi-region writes for global distribution
- D
Use Azure Synapse Analytics as the primary data store for both reads and writes
Show answer and explanation
Correct answer: A
Explanation
To handle sudden spikes in read-heavy traffic with minimal changes, enabling read-scale replicas in the Business Critical or Premium tiers of Azure SQL Database is a recommended approach. This architecture allows read-only replicas to automatically offload queries from the primary, ensuring better performance during peak loads. For reference, see Microsoft documentation on Azure SQL Database read-scale replicas (https://learn.microsoft.com/azure/azure-sql/database/read-scale-out).
- A. Correct.
Correct: Upgrading to the Business Critical tier provides built-in read-scale replicas for Azure SQL Database. This offloads read queries to secondary replicas, increasing throughput without substantial code modifications. According to Microsoft’s documentation, read-scale replicas automatically handle read workloads, making it a straightforward approach to scaling read performance.
- B. Incorrect.
Incorrect: While manual sharding can distribute data, it requires significant architectural changes and adds complexity. It is not necessarily minimal in terms of code and operational overhead, especially if the existing application code is not designed for sharding.
- C. Incorrect.
Incorrect: Azure Cosmos DB does offer global distribution and can scale massively, but migrating an existing relational workload might require major schema and query adjustments. This does not meet the requirement for minimal application changes.
- D. Incorrect.
Incorrect: Azure Synapse Analytics is primarily designed for large-scale data warehousing and analytics. Using Synapse as the main OLTP store for frequent read/write transactions is impractical, and it would not directly solve the high read concurrency in a transactional workload.