MCQ Flashcards

A company needs to connect their on-premises data center to Google Cloud with 10 Gbps bandwidth, consistent latency, and an SLA. Which option is correct?

Answer: Dedicated Interconnect with 10 Gbps circuit

Dedicated Interconnect provides a direct physical connection to Google with a 99.99% SLA (with redundant circuits) and guaranteed bandwidth without traversing the public internet.

An architect needs to implement a blue/green deployment for a Cloud Run service to allow instant rollback. What is the correct mechanism?

Answer: Use Cloud Run traffic splitting to route 100% traffic to the new revision, retaining the old revision

Cloud Run revision traffic splitting allows atomic traffic shifts between revisions, and rolling back simply redirects 100% traffic to the prior revision instantly.

A healthcare application must ensure that data never leaves a specific geographic region for compliance. Which combination of controls enforces this?

Answer: Use VPC Service Controls with a perimeter and resource location restrictions via Organization Policy

VPC Service Controls restrict API access to within the perimeter, and the constraints/gcp.resourceLocations policy prevents resource creation outside approved regions.

A team wants to use Terraform to manage Google Cloud infrastructure and prevent drift between the desired and actual state. What workflow best achieves this?

Answer: Store Terraform state in a Cloud Storage bucket, use a CI/CD pipeline to run terraform plan on PRs and terraform apply on merge

Remote state in Cloud Storage with CI/CD-enforced plan-then-apply workflow ensures state is shared, changes are reviewed, and applies are auditable.

Your BigQuery costs are high due to analysts running full table scans on a large table frequently queried by date. What is the most cost-effective optimization?

Answer: Partition the table by date and require analysts to include the partition filter in queries

Date partitioning restricts bytes scanned to only relevant partitions when a date filter is applied, directly reducing per-query cost.

A team needs to run a GPU-accelerated ML training job that takes 8 hours and can tolerate restarts. How should they minimize cost?

Answer: Use Spot (preemptible) A2 GPU VMs with checkpointing enabled

Spot/preemptible GPU VMs cost up to 60-91% less than on-demand; checkpointing ensures restarts don't lose progress, making them viable for long fault-tolerant training jobs.

A professional services firm needs to give clients temporary read access to specific BigQuery datasets that expires automatically after 30 days. What is the best approach?

Answer: Use IAM Conditions with a date expression to grant time-bound access to the dataset

IAM Conditions allow you to attach a time-based expression (request.time < timestamp) to any role binding, automatically making it invalid after the specified date.