GCP Professional Cloud Architect Practice Question

A ride-sharing company plans to migrate its global trip-matching platform from an on-premises MySQL cluster to Google Cloud. The service must process around 20 000 writes and 30 000 reads per second from riders and drivers worldwide, guarantee serializable transaction consistency, offer at least 99.999 % availability across regions, automatically scale without manual sharding, and expose a familiar SQL interface to developers. Which managed data service best satisfies these business requirements?

Firestore in multi-region (Datastore) mode
Cloud Spanner with a multi-region instance
Cloud Bigtable configured with multi-cluster replication
Cloud SQL for MySQL using cross-region read replicas

Report Issue

Answer Description

Cloud Spanner is engineered for globally distributed, horizontally scalable OLTP workloads. It delivers strong, serializable consistency across regions, provides a 99.999 % multi-region SLA, and supports ANSI-SQL, removing the need for manual sharding or complex replication logic.

Cloud Bigtable can replicate data across regions but offers eventual consistency for multi-cluster writes and lacks a relational SQL interface. Cloud SQL for MySQL is limited to regional primary instances with optional read replicas and offers a lower availability SLA; it cannot provide multi-region active-active writes or seamless horizontal scaling. Firestore (in either Native or Datastore mode) supplies automatic scaling but is NoSQL, not relational SQL, and imposes different consistency and transactional limits. Therefore, Cloud Spanner is the only option that meets every stated business requirement.

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.