GCP Professional Cloud Architect Practice Question

Your company runs hundreds of Compute Engine VMs in both us-central1 and europe-west1. Each VM has multiple 500 GB Balanced Persistent Disks that store order-processing data. A new compliance mandate states:

Every disk must be backed up at least once every 6 hours.
Backups must reside outside the source disk's region so they survive a regional outage.
Storage costs must be minimized by avoiding duplicate copies of unchanged blocks, even when different disks share common data.
Any newly created disks on existing or future VMs must follow the same backup policy without requiring new scripts.
Administrators must be able to restore the most recent backup directly to a new Persistent Disk in the same project within minutes.

Which approach best meets all requirements while keeping operational overhead and costs low?

Create a Compute Engine snapshot schedule resource policy that takes snapshots every 6 hours and stores them in a different region. Attach the policy to all existing disk resources and set it as the default for new disks. Rely on incremental Persistent Disk snapshots for cross-disk deduplication and fast, in-place restores.
Enable synchronous block-level replication to regional Persistent Disks in a second region and manually trigger a full snapshot once a week.
Use Cloud Scheduler to invoke a Cloud Function every 6 hours that runs gcloud compute disks snapshot for each disk, then copies the snapshot files to a Cloud Storage bucket in another region.
Install backup agents inside each VM that archive filesystem data to Cloud Storage every 6 hours; restore by recreating disks from the archived files when needed.

Report Issue

Answer Description

Using a Compute Engine snapshot schedule policy meets all stated needs: you can create a resource policy that takes snapshots every 6 hours and sets the snapshot storage location to a secondary region (or multi-region) different from the source disks' region, satisfying the geographic-isolation requirement. When applied to an entire project, folder, or disk-creating instance template, the policy is automatically inherited by any future disks, eliminating manual scripting. Persistent Disk snapshots are incremental at the block level and globally deduplicated across disks that share identical data, so only changed blocks are stored, which keeps storage costs low. During recovery you can create a new Persistent Disk in minutes from the latest snapshot using the console, gcloud, or REST-fulfilling the RTO goal. Alternatives require custom automation (manual gcloud + copy), store full data sets without cross-disk deduplication, or rely on unsupported cross-region disk replication or guest-level backups that lengthen recovery and add operational burden.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.