PROFESSIONAL-CLOUD-DEVELOPER · Question #154
You recently developed a new service on Cloud Run. The new service authenticates using a custom service and then writes transactional information to a Cloud Spanner database. You need to verify that…
The correct answer is B. Create a Google Kubernetes Engine cluster running the Locust or JMeter images to dynamically. GKE running Locust or JMeter images is the correct choice because both are industry-standard distributed load testing tools capable of simulating thousands of concurrent transactions per second. GKE's Horizontal Pod Autoscaler and cluster autoscaler allow the test…
Question
You recently developed a new service on Cloud Run. The new service authenticates using a custom service and then writes transactional information to a Cloud Spanner database. You need to verify that your application can support up to 5,000 read and 1,000 write transactions per second while identifying any bottlenecks that occur. Your test infrastructure must be able to autoscale. What should you do?
Options
- ABuild a test harness to generate requests and deploy it to Cloud Run. Analyze the VPC Flow Logs
- BCreate a Google Kubernetes Engine cluster running the Locust or JMeter images to dynamically
- CCreate a Cloud Task to generate a test load. Use Cloud Scheduler to run 60,000 Cloud Task
- DCreate a Compute Engine instance that uses a LAMP stack image from the Marketplace, and use
How the community answered
(25 responses)- A8% (2)
- B72% (18)
- C4% (1)
- D16% (4)
Explanation
GKE running Locust or JMeter images is the correct choice because both are industry-standard distributed load testing tools capable of simulating thousands of concurrent transactions per second. GKE's Horizontal Pod Autoscaler and cluster autoscaler allow the test infrastructure itself to scale out automatically to generate the required load, satisfying the autoscale requirement. Locust/JMeter also provide detailed metrics to identify bottlenecks. Option A is partially valid (Cloud Run autoscales) but VPC Flow Logs only show network-level data - not application or database bottlenecks. Option C (Cloud Tasks + Cloud Scheduler) is designed for task queuing, not high-frequency load testing. Option D (single Compute Engine LAMP instance) cannot autoscale and won't generate sufficient load.
Topics
Community Discussion
No community discussion yet for this question.