DEA-C01 Practice Question 95: S3 Express One Zone Low Latency
AWS Certified Data Engineer – Associate (DEA-C01) · Test 10 · question 5 of 10
A machine learning training job reads millions of small objects with a random access pattern, and GPU instances sit idle while they wait for data from Amazon S3. The team needs consistent single-digit millisecond object access. The training data can be regenerated from a source dataset, so losing an Availability Zone is acceptable. Which storage choice fits best?
Show answer and explanation
Answer:
S3 Express One Zone is *"a high-performance, single-zone Amazon S3 storage class that is purpose-built to deliver consistent, single-digit millisecond data access for your most latency-sensitive applications,"* and it *"benefits analytics and machine learning workloads ... especially workloads with lots of smaller accesses or large numbers of random accesses."* Objects live in directory buckets, and AWS recommends specifying an Availability Zone *"that's local to your Amazon EC2, Amazon Elastic Kubernetes Service, or Amazon Elastic Container Service (Amazon ECS) compute instances"*. The data is not stored redundantly across Availability Zones, which is acceptable here because it can be regenerated (option A).
- B is wrong because S3 One Zone-IA also uses a single zone, but it is an infrequent-access class with S3 Standard-like first-byte latency, and it charges retrieval fees. It is not built for single-digit millisecond access. - C is wrong because Transfer Acceleration speeds up long-distance transfers over the internet. It does not lower object access latency for compute in the same Region. - D is wrong because Intelligent-Tiering optimizes the cost of data with changing access patterns, not request latency.