GCP Professional Cloud Architect Practice Question

Your team is running a proof-of-concept to migrate a 10 TB on-premises MySQL database to Cloud SQL for MySQL by using Database Migration Service (DMS). Compliance requires an automated report confirming that every target table's row counts and column-level checksums match the source before cut-over. You already have a Cloud Build pipeline that orchestrates the DMS jobs for both staging and production. Which approach best satisfies the validation requirement while maximizing reuse of the existing CI/CD tooling?

Add a post-replication step to Cloud Build that builds (or pulls) a custom Docker image containing the open-source Data Validation Tool, runs it to compare source and target MySQL row counts and column-level checksums, and fails the build if mismatches exceed the configured threshold.
Invoke Cloud Data Loss Prevention (DLP) API scans on both databases from Cloud Build and compare the number of sensitive-data findings to confirm that the two datasets are identical.
Configure Database Migration Service to generate a built-in data consistency report that automatically blocks cut-over if mismatches are found, and have Cloud Deploy proceed only after the report passes.
Trigger a Cloud Function that exports full SQL dumps from both databases, computes whole-file checksums with Cloud Storage, and requires manual review of the checksum values before proceeding.

Report Issue

Answer Description

The open-source Data Validation Tool (DVT) from Google is purpose-built to compare data between source and target stores supported by Database Migration Service, including on-premises MySQL and Cloud SQL for MySQL. DVT can compute row counts and column-level checksums and exits with a non-zero status when discrepancies exceed configured thresholds, allowing Cloud Build to fail the pipeline automatically. Although Google does not publish an official container image, the DVT repository includes a Dockerfile; you can build a custom Docker image in Cloud Build or invoke the CLI directly in a Cloud Build step. This integrates validation into the existing pipeline for both staging and production without introducing unrelated services or relying on non-existent DMS features. Cloud DLP focuses on sensitive-data discovery, SQL dump comparisons require manual effort, and DMS does not generate automatic consistency reports that can gate cut-over.

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.