5.1 Collecting Performance Information in Google Cloud PROFESSIONAL-CLOUD-DEVOPS-ENGINEER Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   Your organization uses Cloud CDN to serve content. Active Assist recommends increasing the cache TTL from 1 hour to 24 hours to reduce origin load and costs. The content is updated daily, and users expect fresh content. What should you do?

   • AReject the recommendation because it will increase latency
   • BApply the recommendation and use cache invalidation for updates
   • CReject the recommendation because content freshness is critical
   • DApply the recommendation to reduce costs
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   Your organization uses Cloud SQL for MySQL. Active Assist recommends moving to a smaller machine type due to low CPU utilization. The database experiences occasional spikes during end-of-month reporting. What should you do?

   • AResize immediately to save costs
   • BUse a read replica to handle the spikes
   • CResize and rely on automatic scaling
   • DKeep the current size and review the recommendation during a non-peak period
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   Your team has a Kubernetes workload on GKE with a HorizontalPodAutoscaler (HPA) configured. Active Assist recommends adjusting the HPA target CPU utilization from 70% to 50% to improve performance. The workload is not latency-sensitive but you want to avoid over-provisioning. What is the most appropriate action?

   • AApply the recommendation to reduce latency
   • BEvaluate the recommendation against your performance requirements and cost budget
   • CApply the recommendation only if the cluster is underutilized
   • DReject the recommendation because it will increase costs
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   Your organization uses BigQuery for analytics. Active Assist recommends partitioning a large table by date to reduce query costs and improve performance. The table is currently partitioned by ingestion time, but queries frequently filter on a transaction_date column. What should you do?

   • AIgnore the recommendation because partitioning is already enabled
   • BAdd a clustering key on transaction_date instead
   • CCreate a new table partitioned by transaction_date and migrate the data
   • DUse a materialized view to pre-aggregate data
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   You are troubleshooting a performance issue in a distributed application. Cloud Trace shows that a specific service has high latency, but Cloud Monitoring shows low CPU and memory utilization for that service. What is the most likely cause?

   • AThe service is memory-bound
   • BThe service is experiencing a network bottleneck
   • CThe service is CPU-bound
   • DThe service is waiting on an external dependency
   Reveal Solution Discussion Select an answer first