home / exams / Google / gcp-ace / Practice Test 1
gcp-ace Google

Google Associate Cloud Engineer - Practice Test 1

Question #26
Data Analytics

You are planning to execute a complex query in Google BigQuery and anticipate a large result set. Your project uses on-demand pricing, and you need to estimate the cost of running this query before execution. What is the most effective way to achieve this?

A. Temporarily switch your BigQuery project to Flat-Rate pricing for the query, then revert to on-demand pricing.
B. Utilize the command line to perform a dry run of the query to estimate the bytes processed (bytes read), then use the Google Cloud Pricing Calculator to convert this estimate to a dollar amount.
C. Execute a dry run of the query via the command line to estimate the number of bytes returned, then use the Google Cloud Pricing Calculator to determine the cost.
D. Run a `SELECT COUNT(*)` query to approximate the number of rows the main query will scan, and then use the Google Cloud Pricing Calculator to estimate the cost based on this row count.
Question #27
IAM & Security

Your organization's DevOps team requires comprehensive control over Compute Engine resources within a specific development project. However, they must be restricted from creating or modifying any other resource types in that project. Following Google's recommended practices for permission management, what is the most appropriate action to take?

A. Grant the basic role `roles/viewer` and the predefined role `roles/compute.admin` to the DevOps group.
B. Create an IAM policy and grant all `compute.instanceAdmin.*` permissions to the policy, then attach the policy to the DevOps group.
C. Create a custom role at the folder level, grant all `compute.instanceAdmin.*` permissions to it, and then assign this custom role to the DevOps group.
D. Grant the basic role `roles/editor` to the DevOps group.
Question #28
Serverless

Your development team is preparing to deploy a new feature to an existing Cloud Run service in production. To minimize the impact of potential issues and adhere to Google's recommended practices for managing service revisions, you want to reduce the number of users affected by any outage without incurring additional development or operational costs for your customers. What is the most appropriate strategy?

A. Advise customers to implement exponential backoff for retrying service access to mitigate problems after the new revision deployment.
B. Implement a gradual rollout of the new revision, splitting customer traffic between the old and new revisions to enable quick rollback if issues arise.
C. Direct all customer traffic to the new revision immediately, and revert to the previous revision if any problems are observed in production.
D. Deploy the application to a separate, new Cloud Run service and instruct customers to switch to using this new service.
Question #29
Deployment Management

You have made significant modifications to a complex Google Cloud Deployment Manager template and need to quickly verify that all defined resources have their dependencies correctly satisfied before committing the changes. Which action provides the most rapid feedback on your modifications?

A. Implement granular logging statements within a Deployment Manager template authored in Python.
B. Execute the Deployment Manager template using the "--preview" option within the same project, and examine the state of interdependent resources.
C. Monitor the activity of the Deployment Manager execution on the Cloud Logging page of the Google Cloud console.
D. Execute the Deployment Manager template against a separate project with an identical configuration, and monitor for failures.
Question #30
Networking

Your company operates in a hybrid cloud environment, with some applications deployed on Google Cloud and others on-premises. A VPN tunnel connects your Google Cloud VPC to your on-premises network. Several applications in Google Cloud need to access an on-premises database server. To ensure that applications can find the database even if its IP address changes, without requiring code modifications in each application, what is the most appropriate solution?

A. Configure Cloud NAT for all subnets of your VPC to be used when egressing from the VM instances.
B. Create a private zone on Cloud DNS, and configure the applications to use a DNS name for the database.
C. Configure the IP of the database as custom metadata for each instance, and query the metadata server.
D. Query the Compute Engine internal DNS from the applications to retrieve the IP of the database.
Back to gcp-ace 50 questions - page 6 of 10