GCP Professional Data Engineer Practice Question

Your team is re-platforming a regional on-premises PostgreSQL database that handles thousands of OLTP transactions per minute for an e-commerce web application. Key requirements are:

Remain fully PostgreSQL-compatible to avoid code changes.
Provide automatic backups, point-in-time recovery, and automated patching.
Achieve at least 99.95 % availability within a single Google Cloud region and support creating optional read replicas for analytics.
Minimize operational overhead and total cost; global distribution and unlimited horizontal write scaling are not required.

Which managed Google Cloud service best satisfies these requirements?

Cloud Spanner with a single-region instance
Cloud SQL for PostgreSQL with high-availability enabled and optional read replicas
Cloud Bigtable with a single-region multi-cluster instance
AlloyDB for PostgreSQL with a regional primary and read pools

Report Issue

Answer Description

Cloud SQL is Google Cloud's fully managed service for MySQL, PostgreSQL, and SQL Server. It offers native PostgreSQL engines, so the application can migrate with no code changes. When deployed in a high-availability configuration, Cloud SQL provides 99.95 % regional availability, automatic backups, point-in-time recovery, and managed maintenance. It also supports in-region and cross-region read replicas for reporting workloads. Spanner delivers higher availability and global horizontal scaling but uses a different dialect and costs more than needed for this regional workload. Bigtable is a NoSQL wide-column store lacking PostgreSQL compatibility and ACID transactions. AlloyDB for PostgreSQL offers enhanced performance but targets demanding, larger-scale workloads at a higher price point; Cloud SQL remains the most cost-effective fit for standard OLTP with the stated SLAs and feature needs.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.