The correct answer is Host the visualization tool in the same AWS Region as the data warehouse and access it over a Direct Connect connection at a location in the same Region.

This setup ensures that large query results (50 MB per request) are transferred internally within the same Region in AWS, avoiding internet-based data transfer charges. Then, only the small, processed visualization output (~500 KB) needs to traverse the Direct Connect link to on-premises users. Since Direct Connect offers lower and more predictable egress pricing than standard internet data transfer, and same-Region Direct Connect transfers are especially cost-effective, this setup provides the lowest data transfer costs overall.

Host the visualization tool on premises and query the data warehouse directly over the internet is the most expensive option. Data transfer out to the internet from AWS incurs the highest rates, especially for large payloads like the 50 MB query results being returned.

Host the visualization tool in the same AWS Region as the data warehouse. Access it over the internet avoids AWS-internal transfer costs between the tool and the data warehouse but still sends the large 50 MB result sets to on-premises users via the public internetagain incurring higher egress fees.

Host the visualization tool on premises and query the data warehouse directly over a Direct Connect connection at a location in the same AWS Region improves cost compared to internet-based transfer, but still requires sending the full 50 MB query responses over Direct Connect. While cheaper than internet, this is still more costly than sending only the 500 KB post-processed visualization output if the tool is hosted in-region.

Why this question matters: It tests your ability to reduce operational costs by strategically placing compute resources to minimize expensive data transfers. Understanding how data movesespecially across AWS services, Regions, Direct Connect links, and the internetis critical for cost-efficient cloud design.

Tip: When AWS Direct Connect is involved, always ask: where is the bulk of the data traveling? Hosting compute close to where large data lives can lead to substantial cost savings.

Full AWS Practitioner Certification Question