Question

Difficulty: MediumDesigning Migration Plans and Data Transfer Strategies

An autonomous mobility platform needs to migrate 40 TB of unstructured sensor logs from an on-premises storage system and a 500 GB relational database to Google Cloud within a 3-week planning window. The organization maintains an active 1 Gbps Dedicated Interconnect with 600 Mbps of consistent available bandwidth allocated for data migration. The operational requirement dictates that the relational database must remain fully operational for online transactions until a planned 2-hour final cutover window.

Which TWO architectural recommendations should be included in the cloud migration strategy to fulfill these business and technical requirements?

  1. Utilize Storage Transfer Service for agent-based transfers to move the 40 TB unstructured log dataset directly over the Dedicated Interconnect connection.Answer
  2. Deploy Database Migration Service (DMS) to execute initial data loading and continuous Change Data Capture (CDC) replication for the relational database prior to final cutover.Answer
  3. C
    Order a Google Cloud Transfer Appliance to perform an offline bulk migration of the 40 TB unstructured log files.
  4. D
    Establish an IPsec High Availability (HA) VPN connection over the public internet specifically to isolate and transmit the database and file data transfers.

Answer

The migration strategy should combine using Storage Transfer Service over the existing Dedicated Interconnect for the 40 TB log files and Database Migration Service with continuous replication for the relational database.
Combining Storage Transfer Service for unstructured logs over the Dedicated Interconnect and Database Migration Service with CDC for relational data satisfies both time constraints and low-downtime operational requirements. Online network migration of 40 TB completes well under a week over 600 Mbps available bandwidth, while continuous CDC replication ensures the database stays synchronized until the 2-hour cutover window.

Step-by-Step Solution

1
Calculate network throughput for the 40 TB file transfer
At 600 Mbps600\text{ Mbps}, transferring 40 TB40\text{ TB} (320,000 Gb320,000\text{ Gb}) requires 320,000/0.6=533,333 seconds6.17 days\approx 320,000 / 0.6 = 533,333\text{ seconds} \approx 6.17\text{ days} of continuous transfer (or under 4 days at full line rate without protocol overhead), which comfortably fits inside the 3-week window.
Determining network capacity confirms whether an online transfer tool is viable versus an offline appliance.
2
Select the appropriate file migration tool
Storage Transfer Service allows automated, scalable, agent-based ingestion of on-premises file data directly into Cloud Storage over private Interconnect connectivity.
Using existing private bandwidth avoids physical device logistics and simplifies orchestration.
3
Select the database migration pattern
Database Migration Service (DMS) provides minimal-downtime database migration using continuous Change Data Capture (CDC).
Continuous replication keeps target database updated in real-time, allowing cutover to take place within the narrow 2-hour window.

Key Concept

Selecting online file transfer mechanisms versus offline appliances based on bandwidth calculations, alongside minimal-downtime database replication strategies.
Rate this question