AZ-305 Design data storage solutions Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Use the export/import bacpac method.

  Why it's wrong here

  The BACPAC export/import method is offline by design: the export process captures a logical schema and data snapshot to a .bacpac file, and any transactions committed after the export starts are not included unless the source database is quiesced or taken offline. For a 500 GB database, the export and subsequent import can take many hours, and the application must be stopped during the entire cutover window to guarantee data consistency, making it incompatible with a minimal-downtime requirement.

• ✗

  Use Azure Data Factory to copy data.

  Why it's wrong here

  Azure Data Factory (ADF) is an ETL/ELT orchestration service that performs one-time or scheduled bulk copies between disparate data stores, but it does not provide continuous log-based change replication or a managed cutover mechanism. While you could script periodic incremental loads, this process is brittle, complex, and introduces latency on the order of the refresh schedule; it also does not automatically migrate database objects such as stored procedures, indexes, and security policies, and it requires the source to remain online without a clean synchronization point, making it unsuitable for a single 500 GB database with strict downtime limits.

• ✗

  Use transactional replication.

  Why it's wrong here

  Transactional replication is not ideal here because it necessitates a full initial data load followed by continuous transaction log shipping, which can introduce significant latency and complexity for a 500 GB database requiring minimal downtime. It is tempting as it offers near real-time data synchronisation and is excellent for ongoing replication scenarios or distributing data across multiple servers, but it is not the most efficient for a single, large, one-time migration with strict downtime constraints.

• ✓

  Use Azure Database Migration Service with online migration.

  Why this is correct

  Azure Database Migration Service with online migration is the correct choice because it performs an initial full copy of the schema and data, then uses a continuous replication stream to keep the target synchronized with the source as transactions occur. It supports a controlled cutover where you can validate the target, stop writes on the source, and switch applications with near-zero downtime. DMS automates the heavy lifting, handles large databases like 500 GB, and is designed specifically for minimal-downtime migrations to Azure SQL, eliminating the need for manual log shipping or custom replication code.

Go deeper

Related to this question