AWS Certified Solutions Architect - Associate / Question #1732 of 1019

Question #1732

A company operates a video streaming service in the us-east-1 Region. The service allows users globally to upload and stream videos, with uploads capped at 500 MB per video. Popular videos are streamed frequently for several weeks, while others lose popularity within days. The application leverages video metadata to recommend content to users. Which solution provides the MOST cost-effective user access?

A

Store the videos in Amazon DynamoDB. Enable DynamoDB Accelerator (DAX) to cache frequently streamed items.

B

Store the videos in the Amazon S3 Intelligent-Tiering storage class. Store the video metadata and its S3 location in DynamoDB.

C

Store the videos in the Amazon S3 Standard storage class. Configure an S3 Lifecycle policy to transition videos older than 14 days to S3 Standard-Infrequent Access (S3 Standard-IA). Use object tags to manage metadata.

D

Store the videos in the Amazon S3 Glacier storage class. Implement an S3 Lifecycle policy to move videos older than 14 days to S3 Glacier Deep Archive. Store the video metadata and its S3 location in Amazon OpenSearch Service.

Explanation

Option B is correct because:
1. S3 Intelligent-Tiering automatically transitions objects between Frequent and Infrequent Access tiers based on access patterns, reducing costs for videos that lose popularity quickly while keeping frequently accessed videos readily available.
2. DynamoDB efficiently stores video metadata and S3 locations, enabling fast queries for recommendations.

Other options are incorrect because:
- A: DynamoDB has a 400KB item limit, making it unsuitable for 500MB videos.
- C: Transitioning to S3 Standard-IA after 14 days risks higher retrieval costs if videos remain popular beyond this period.
- D: S3 Glacier/Deep Archive is designed for archival storage with retrieval delays, which is incompatible with streaming requirements.

Key Points: Use S3 Intelligent-Tiering for variable access patterns and DynamoDB for metadata. Avoid Glacier for streaming and DynamoDB for large files.

Answer

The correct answer is: B