MLA-C01 Practice Question 43: S3 Transfer Acceleration
AWS Certified Machine Learning Engineer – Associate (MLA-C01) · Test 5 · question 3 of 10
A medical imaging startup trains its models from a single Amazon S3 bucket in us-east-1. Partner hospitals in Australia, Brazil, and Germany upload 5-20 GB batches of scans to the bucket every day over the public internet, and they already use multipart uploads. Uploads from the distant sites take hours and often slow the training schedule. The startup wants faster uploads over these long distances without changing where the data is stored or asking hospitals to install hardware.
What should the startup do?
Show answer and explanation
Answer:
S3 Transfer Acceleration is a bucket-level feature for fast transfers over long distances between clients and an S3 bucket. Clients send data to the nearest Amazon CloudFront edge location, and from there it travels to S3 over an optimized AWS network path instead of the public internet the whole way. It suits exactly this pattern of regularly moving gigabytes across continents. It needs only the bucket setting plus the bucket-name.s3-accelerate.amazonaws.com endpoint, and it works alongside multipart uploads (option B).
- A is wrong because replication copies objects *after* they land in a source bucket. Each hospital would upload to a nearer Region, then wait for replication, and the startup would now run and pay for three extra buckets, adding storage locations rather than speeding up the path. - C is wrong because Snowball Edge is physical hardware shipped to a site, which the startup ruled out, and a round trip of days is slower than a daily network upload. - D is wrong because the setup described speeds up *downloads* to viewers. The problem here is uploads into the bucket, and Transfer Acceleration is the S3 feature that uses the edge network for that direction.
Reference: Configuring fast, secure file transfers using Amazon S3 Transfer Acceleration