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 mission-critical, high-availability Cloud SQL for PostgreSQL instance that experienced an unexpected failover event and a spike in query latency overnight. During your post-incident investigation, you need to accomplish the following goals:
Which combination of Google Cloud tools and logs should you analyze?
Combining Cloud Audit Logs, instance database logs in Cloud Logging, and System Insights and Query Insights in Cloud Monitoring provides end-to-end database observability across infrastructure, control-plane, and SQL execution layers. Rather than examining isolated symptoms, this integrated approach links automated failover triggers and administrative changes directly to engine-level errors and query performance spikes.
resource.type="cloudsql_database" and centering the time-range selector on the incident timestamp surfaces PostgreSQL engine logs (capturing lock waits, checkpoint activity, and error statements) right when the disruption occurred.Investigating a high-availability failover alongside a performance degradation requires visibility across both Google Cloud's management plane and the database engine itself. Relying solely on console operation history, network flow samples, or future-facing query flags misses critical historical context. Correlating Cloud Audit Logs, timestamp-centered Cloud Logging entries, System Insights, and Query Insights directly answers who or what triggered the failover and which workload patterns drove the latency spike.
Admin Activity audit logs record user-driven API calls that modify Google Cloud resource configurations. VPC Flow Logs capture sampled network traffic flows sent from and received by VM instances within a VPC subnet. Cloud SQL recommenders analyze historical resource usage using heuristics and machine learning to suggest optimizations such as resizing instances or preventing out-of-disk failures.
You wanted to audit whether an engineer manually modified an instance configuration via API, check sampled subnet-level network connectivity, and evaluate long-term instance rightsizing recommendations.
While Cloud SQL recommenders can highlight chronic underprovisioning that increases failover risk, post-incident forensics requires System Event audit logs, Logs Explorer, System Insights, and Query Insights to pinpoint exact causes.
The Operations and logs pane on a Cloud SQL instance Overview page lists high-level administrative tasks performed on the instance, including operation type, completion time, status, and outcome messages. Data Access audit logs record authenticated, user-driven API calls that read resource metadata or read and write user-provided data.
gcloud CLI, or Cloud SQL Admin API (such as user management or instance edits). It does not capture database-level errors, lock contention, or internal engine diagnostics.You needed to verify the completion status of a recent administrative task initiated through the gcloud CLI (such as a password change or instance resize) and wanted to start auditing future API-based metadata reads for compliance.
A common misconception is that enabling Data Access audit logs logs all SQL queries inside PostgreSQL. In practice, SQL query diagnostics and execution plans belong in Query Insights or targeted database logging flags, while infrastructure failovers are tracked in System Event audit logs.
In PostgreSQL, log_statement controls which SQL statements are written to the database logs, and logging_collector manages capturing stderr output into log files. Cloud Trace is a distributed tracing service that collects request latency distributions and execution spans across instrumented application components.
log_statement and logging_collector only affects future logging behavior and does not expose historical Data Access audit logs or overnight failover records.log_statement and logging_collector consumes significant CPU and memory on a Cloud SQL instance, which can worsen latency on a heavily loaded production database.log_statement to none and logging_collector to off specifically to prevent logging from saturating CPU and memory.Your goal was to capture future SQL statements for short-term debugging in a non-critical test environment while separately profiling end-to-end HTTP request latency across instrumented microservices.
For deep-dive PostgreSQL diagnostics without excessive overhead, Query Insights collects query telemetry via built-in in-memory agents, while targeted flags like log_lock_waits and log_min_error_statement capture lock contention more efficiently than logging every statement.
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