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.
A retail company is migrating a 10 TB transactional MySQL database from an on-premises data center to Spanner on Google Cloud over a 10 Gbps dedicated interconnect. The database handles 24/7 customer orders, and the migration must meet the following requirements:
Which migration strategy and toolset should the database engineer implement?
Storage Transfer Service is a managed service for transferring large volumes of file and object data from on-premises file systems or other cloud storage providers into Cloud Storage over HTTPS using lightweight on-premises agents.
The MySQL database were relatively static or the business could afford a multi-hour scheduled maintenance window during off-peak hours to halt all writes, export the database, transfer the dump files, and load them into the target system.
For large databases, engineers sometimes use batch dump-and-load tools only for the initial seed, but achieving RPO = 0 with sub-5-minute downtime still requires pairing the seed with continuous replication from a recorded log position.
The Y (writing and reading) pattern is a parallel migration strategy in which applications are modified to write every new transaction simultaneously to both the legacy source database and the new target database while continuing to read from the legacy database until backfill and validation finish.
The database had zero tolerance for even a brief cutover pause (requiring true zero-downtime parallel adoption), the source engine did not support log-based CDC, and the engineering team had the time and budget to refactor application code.
When zero-refactoring is prioritized alongside minimal downtime, continuous replication using transaction log-based CDC achieves near-zero cutover downtime at the infrastructure level without altering application code.
Transfer Appliance is a high-capacity, hardware-based offline data transfer device that you connect to an on-premises network, load with data, and physically ship to a Google Cloud data center for ingestion into Cloud Storage.
The organization needed to migrate hundreds of terabytes or petabytes of static or archival data from a location with very slow network bandwidth (such as under 100 Mbps) and could tolerate either days of downtime or an offline seed followed by a separate CDC phase.
Even when physical appliances are used to seed massive datasets, mission-critical transactional databases with strict RTO and RPO targets still require an online continuous replication tool like Datastream to synchronize changes generated during transit.
Datastream is a serverless Change Data Capture (CDC) and replication service that extracts initial historical snapshots (backfill) and streams ongoing transaction log changes in near real time. When paired with Dataflow streaming templates, Datastream writes CDC events from Cloud Storage directly into Spanner to maintain a continuously synchronized target database.
INSERT, UPDATE, and DELETE in sequence. Waiting for Datastream event counts and Dataflow replication lag metrics to drop to zero before cutover guarantees complete data consistency.Unlike offline bulk transfers that require hours of downtime to freeze writes and move 10 TB, continuous CDC replication decouples the bulk data transfer from the cutover window. It achieves near-zero downtime and zero data loss without the development cost and split-brain risks of application-level dual writes.
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