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 manages a production Cloud SQL for PostgreSQL regional instance configured with high availability (HA) across us-central1-a (primary) and us-central1-b (standby), along with an in-region read replica in us-central1-c. Before a major release, you need to execute a failover drill to verify standby readiness and evaluate its operational impact on:
Which execution plan and expected outcome accurately describes this HA failover drill?
While taking backups before risky changes is a common operational habit, combining an active backup with a Cloud SQL HA failover drill violates Cloud SQL's failover preconditions and misrepresents how read replicas and HA standby instances behave.
us-central1-c), never to the HA standby node.You waited for any backup operations to finish completely before initiating failover, expected read replica connections to drop and reconnect alongside primary connections, and used a dedicated read replica rather than the HA standby to offload read traffic.
Read replica promotion converts an asynchronous read replica into an independent primary database instance, permanently severing its replication link from the original primary. While useful for cross-region disaster recovery or splitting workloads, it does not test the high availability (HA) standby instance located in us-central1-b.
You were recovering a standalone (non-HA) instance from a zonal outage or executing a cross-region disaster recovery exercise using a read replica, and you planned to reconfigure application connection strings to target the promoted instance's new IP address.
This approach assumes that a Cloud SQL HA standby instance exposes a separate client-facing IP address and that the cluster automatically fails back to the original primary zone once recovery completes. Both assumptions contradict how Cloud SQL HA networking and post-failover role management work.
us-central1-b assumes the primary role, it continues serving as the primary instance even after us-central1-a comes back online.us-central1-a requires intentionally triggering a second manual failover (failback) during a planned window.You were restoring a standalone instance via point-in-time recovery (which creates a new instance with a different IP address requiring client updates) and intentionally scheduling a separate manual failback procedure rather than expecting automatic failback.
In Cloud SQL for PostgreSQL, a regional high availability (HA) instance pairs a primary instance in one zone with a standby instance in a secondary zone using synchronous replication to regional persistent disks. Initiating a manual failover intentionally switches traffic to the standby instance, making it an ideal drill for validating standby readiness and analyzing application resilience.
us-central1-b.us-central1-a is healthy.us-central1-c) automatically reconnect to the new primary once existing connections reset.Unlike promoting a read replica or reconfiguring endpoints manually, initiating a manual HA failover directly exercises the production standby failover path, accurately measures the ~60-second reconnection window, and confirms that clients reconnect seamlessly via the shared static IP address.
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