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 runs a high-availability Cloud SQL for PostgreSQL deployment with a primary instance, a regional standby instance, and two read replicas. During recent peak traffic events, the database team needed a centralized view to diagnose emerging availability risks, including:
Which monitoring strategy should you implement to visualize these critical metrics across your database topology?
Exporting database operational logs (such as postgres.log) to BigQuery creates a long-term analytical archive where engineers can run SQL queries over historical event records.
The requirement were to retain historical audit logs for multi-quarter capacity forecasting or forensic compliance reviews rather than real-time high-availability dashboard visualization.
When monitoring user-facing and high-availability systems, tracking metric distributions (like the 99th percentile) in real time reveals imbalances that daily averages completely obscure.
The predefined Cloud SQL dashboard in Cloud Monitoring provides curated default charts for general instance health, while pg_stat_replication is a PostgreSQL system catalog view used on a primary node to inspect WAL sender status and replica lag via SQL.
pg_stat_replication are not possible on the standby.pg_stat_replication: Even in self-managed PostgreSQL topologies (such as AlloyDB Omni), pg_stat_replication reports streaming status from the primary instance, not from a standby node.psql adds operational overhead and does not provide integrated Cloud Monitoring time-series visualization.You only needed a quick baseline overview from the default dashboard and were managing a self-hosted PostgreSQL cluster where you inspect pg_stat_replication on the primary node.
The Cloud SQL instance Overview page is a built-in console view for inspecting individual metrics—such as CPU utilization, Active connections, Storage usage, and Replication lag—one metric and one instance at a time via a drop-down chart selector.
You only needed to perform a quick, ad-hoc check of a single metric on one specific read replica or primary instance rather than building a centralized dashboard across the entire database topology.
A common misconception is treating a Cloud SQL HA standby instance like an active read replica. Because the standby only activates during an automated failover (typically after roughly 60 seconds of primary unresponsiveness), replication lag monitoring focuses on asynchronous read replicas.
A custom Cloud Monitoring dashboard is a user-defined visualization workspace that aggregates real-time time-series metrics from multiple Google Cloud resources onto a single screen. Unlike single-instance console views, custom dashboards allow database engineers to select specific metrics, apply aggregations or percentile distributions, and compare primary and read replica behavior side by side.
Cloud SQL automatically streams these core metrics to Cloud Monitoring with no agent installation required. Building a custom dashboard combines the exact high-availability indicators needed into a single pane of glass, overcoming the single-resource limitations of instance overview pages and the lack of specificity in generic predefined dashboards.
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