Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Professional Cloud Database Engineer practice bank has an additional 769 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
Your team is migrating a mission-critical transactional MySQL database to Spanner using Datastream and Dataflow. To safeguard against post-cutover performance degradation, you must meet the following requirements:
Which combination of fallback strategy and data validation framework should you implement?
The duplicate writes (Y) pattern modifies applications or introduces a centralized data-access microservice so that every write operation is sent simultaneously to both the source and target databases during the transition period. DbFit is an open-source database testing framework designed for test-driven development and basic automated regression checks within CI/CD pipelines.
This approach would be appropriate if you had sufficient engineering resources to refactor applications for zero-downtime parallel adoption and only needed lightweight CI/CD integration tests for simple SQL logic.
Reverse replication configures a prebuilt Dataflow pipeline (Spanner_to_SourceDb) powered by Spanner change streams to continuously replicate post-cutover writes from Spanner back to the original MySQL database. The Data Validation Tool (DVT) is a Google-backed, open-source Python CLI framework engineered for automated, multi-level data comparison across heterogeneous environments.
AVG, COUNT, SUM, MIN, MAX, GROUP BY), and row-level (exact match and hash) validations.forward_migration mode) prevents infinite replication loops.Unlike simple fallback or custom scripts, prebuilt reverse replication paired with DVT provides an automated, auditable safeguard that preserves transaction integrity across heterogeneous engines without complicating application code.
A simple fallback strategy reverts applications directly back to the original source database after cutover, either discarding transactions committed on the target database or relying on backups to manually reconstruct lost writes. HammerDB is an open-source database benchmarking and load-testing tool that simulates transactional (TPROC-C) and analytical (TPROC-H) workloads.
A simple fallback combined with HammerDB would fit a migration where extended downtime during a rollback is acceptable, post-cutover transactions can be discarded, and the immediate goal is benchmarking target compute and storage sizing.
Forward replication replicates post-cutover writes from the new target database to a separate, third database instance rather than back to the original source server, allowing applications to execute read-only queries if the primary target becomes unavailable. Custom row-count scripts query COUNT(*) totals across tables to check whether record quantities match.
Forward replication is recommended when you must decommission and discard the original source database immediately after cutover while still retaining an end-to-end tested read-only failover target.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated