Your company wants to migrate their 10-TB on-premises database export into Cloud Storage. You want to minimize the time it takes to complete this activity and the overall cost. The bandwidth between…

Question

Options

• ADevelop a Dataflow job to read data directly from the database and write it into Cloud Storage.
• BUse the Data Transfer appliance to perform an offline migration.
• CUse a commercial partner ETL solution to extract the data from the on-premises database and
• DUpload the data with gcloud storage cp.

• A
  3% (1)
• B
  84% (32)
• C
  11% (4)
• D
  3% (1)

Explanation

Option B is correct because at 1 Gbps of shared network bandwidth, transferring 10 TB over the wire would take roughly 22+ hours under ideal conditions - but real-world factors like latency, overhead, retries, and bandwidth contention make offline migration via the Transfer Appliance significantly faster and more cost-effective for large datasets. Google recommends offline transfer when network throughput makes online transfer impractical or slow at scale, and it eliminates ongoing egress/transfer costs.

Why the others are wrong:

• A (Dataflow): Dataflow is designed for data transformation and processing pipelines, not bulk file migration - it adds unnecessary complexity and cost for a straightforward export-and-load task.
• C (Commercial ETL): A third-party ETL solution introduces additional licensing costs and complexity with no benefit over Google-native tooling for a one-time migration.
• D (gcloud storage cp): While functional for smaller datasets, uploading 10 TB over a 1 Gbps link is slow in practice and subject to network interruptions - not the recommended approach when minimizing time is the goal.

Memory tip: Think "big data, slow pipe = ship the box." When data size is large and bandwidth is limited, Google's offline Transfer Appliance beats the network. If the transfer would take more than a week or the bandwidth is clearly insufficient, go physical.

No community discussion yet for this question.