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 organization is migrating a 2 TB production Amazon RDS for PostgreSQL database supporting a 24/7 transactional application to AlloyDB for PostgreSQL. The migration plan specifies the following requirements:
Which migration strategy should you implement?
A one-time export and import (or backup and restore) strategy uses native database utilities—such as pg_dump and pg_restore—or a one-time dump job to extract a static copy of the source database and load it into the target AlloyDB for PostgreSQL cluster in a single batch operation.
One-time export and import is ideal for small, infrequently used databases or development and reporting workloads where an extended scheduled maintenance window is acceptable and simplicity outweighs availability requirements.
Database Migration Service (DMS) is a fully managed migration service that orchestrates online, homogeneous database migrations to AlloyDB for PostgreSQL. For continuous migrations, DMS connects the destination AlloyDB primary instance to the source PostgreSQL database and manages an automated pipeline consisting of an initial consistent snapshot, schema constraint and index creation, and ongoing Change Data Capture (CDC) via PostgreSQL write-ahead log (WAL) logical decoding.
pglogical extension (or pgnative in quick-start migrations) while the source stays online. During cutover, you briefly pause source writes, wait for replication lag to reach zero, and promote AlloyDB to a standalone primary.pgnative can even migrate into an existing AlloyDB cluster without requiring pglogical installation.Unlike staging CDC streams through Cloud Storage and Dataflow or rewriting application data-access layers, DMS provides a purpose-built, direct-to-destination continuous replication path that minimizes infrastructure overhead and cutover risk for homogeneous PostgreSQL transitions.
Datastream is a serverless, agentless Change Data Capture (CDC) and replication service designed to stream historical backfills and real-time database changes into BigQuery or Cloud Storage. To replicate data into transactional destinations such as AlloyDB for PostgreSQL or Cloud SQL, Datastream writes change events to Cloud Storage, and a downstream Dataflow pipeline (using a Dataflow Flex Template) reads those files and applies the changes to the target database.
This approach would be optimal for heterogeneous migrations (such as migrating MySQL to Spanner), streaming operational databases into BigQuery for analytics, or scenarios requiring complex in-flight data transformations in Dataflow.
Application-level dual writes (often called Y-writing or using a data-access microservice) is a migration pattern in which application code is modified to write every insert, update, and delete operation to both the source database and the target database simultaneously, while gradually shifting read traffic to the new system.
Application-level dual writes are recommended only when zero migration downtime and an immediate, zero-downtime fallback path are strictly mandatory, and the engineering team has the time and budget to refactor applications or build a coordinated data-access microservice.
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