An energy utility company is migrating its smart grid analytics and operations workload to… — Google Cloud Professional Cloud Architect

An energy utility company is migrating its smart grid analytics and operations workload to Google Cloud. The current environment consists of $3.5\text{ PB}$ of historical smart meter reading archives stored on an on-premises Storage Area Network (SAN) and an active $8\text{ TB}$ operational MySQL database used for billing. The enterprise has a $1\text{ Gbps}$ Dedicated Interconnect link with a policy capping cloud migration bandwidth to $300\text{ Mbps}$ . The target architecture requires minimal downtime for the operational database during cutover (under $1\text{ hour}$ ) and complete data integrity across all systems. In what sequence should the Cloud Architect execute these steps to perform the migration?

1Order and configure a Google Cloud Transfer Appliance to perform an offline bulk seed of the $3.5\text{ PB}$ historical SAN archive to Cloud Storage.
2Provision Cloud SQL for MySQL and set up Database Migration Service (DMS) with continuous Change Data Capture (CDC) over the Dedicated Interconnect.
3Run Storage Transfer Service to execute an incremental catch-up sync for files generated on the SAN while the Transfer Appliance was in transit.
4Stop application write traffic to the on-premises MySQL database and monitor DMS to ensure replication lag drops to zero.
5Promote the Cloud SQL database instance to primary, reconfigure application endpoints to target Google Cloud resources, and resume full production workloads.

Answer

The correct migration sequence is: 1) Order and seed bulk historical archives (

3.5\text{ PB}

) offline using Transfer Appliance; 2) Configure Database Migration Service (DMS) with CDC to continuously replicate the active

8\text{ TB}

MySQL database over Dedicated Interconnect; 3) Execute Storage Transfer Service delta sync for incremental updates accrued during shipping; 4) Stop on-premises write traffic and wait for DMS replication lag to reach zero; 5) Promote Cloud SQL to primary and update application connection endpoints.

The correct operational sequence begins by addressing the largest migration bottleneck offline via Transfer Appliance (

3.5\text{ PB}

bulk historical data). Next, continuous online replication for the active database (

8\text{ TB}

MySQL) is established using Database Migration Service with CDC over the Dedicated Interconnect. Once the physical appliance data is loaded in GCP, Storage Transfer Service updates any delta changes accrued during appliance shipping. Finally, write traffic to the source database is stopped during the cutover window to achieve zero replication lag before promoting Cloud SQL and updating application endpoints.

Step-by-Step Solution

Seed bulk offline archive data (

3.5\text{ PB}

) via Transfer Appliance.

Initial

3.5\text{ PB}

archive is safely shipped and loaded into Cloud Storage without choking the

300\text{ Mbps}

network cap.

Petabyte-scale transfers over limited bandwidth require physical appliance offloading to meet project schedules.

Establish continuous online database replication using DMS with CDC for the

8\text{ TB}

MySQL workload.

Database replication runs in the background while production write operations continue uninterrupted on-premises.

Continuous CDC minimizes cutover downtime to just the time needed for final log flush and endpoint redirection.

Perform delta catch-up sync using Storage Transfer Service.

Incremental SAN updates generated during Appliance transit are synced to Cloud Storage.

Guarantees complete file storage sync before cutting over application workflows.

Initiate cutover window by stopping writes on-premises and confirming zero DMS replication lag.

Ensures no un-replicated transactions remain in the pipeline.

Prevents split-brain scenarios and data loss before database promotion.

Promote Cloud SQL to primary and update application configurations.

Application traffic resumes pointing directly to Google Cloud services within the

1\text{-hour}

downtime limit.

Final step to complete workload migration and switch over production operations.

Key Concept

Combining offline bulk transfer (Transfer Appliance) for multi-petabyte historical datasets with online continuous replication (DMS with CDC) for transactional databases to minimize cutover downtime and network saturation.

Estimated Time:2m 30s

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics