You have S3 data with unpredictable access patterns. Frequently accessed and rarely accessed data are mixed together, and you want to optimize storage cost without increasing operational overhead. Which storage class is the most appropriate?

Explanation

S3 Intelligent-Tiering automatically moves objects across tiers based on access patterns. It is well suited to cost optimization for data with unpredictable access frequency.

Per-choice explanations

Practical Use

S3 Intelligent-Tiering monitors the access frequency of each object and automatically moves it across storage tiers (Frequent Access, Infrequent Access, Archive Instant Access, Archive Access, and Deep Archive Access). It delivers cost optimization with no operational overhead for datasets where access patterns are unknown or hard to predict and where frequently and infrequently accessed objects are mixed, such as user uploads, media archives, and log data. There are no retrieval fees; the only additional cost is a minimum object size charge (objects under 128 KB are treated as a single tier) and a small per-object monitoring fee. Compared with manual tier transitions through Lifecycle rules, Intelligent-Tiering is stronger when access patterns cannot be known in advance. Conversely, when access patterns are clear (for example, logs or backups that reliably cool after 30 days), moving directly to Standard-IA or Glacier via Lifecycle is cheaper because no monitoring fee is incurred. Note also that Archive Access and Deep Archive Access require retrieval times ranging from several hours to more than ten hours, so they do not suit frequently accessed data, and this should be reflected in the design. On SAA-C03, conditions such as "unpredictable access patterns," "optimize cost without increasing operational overhead," and "frequent and infrequent access are mixed" point to Intelligent-Tiering. Continuing to use only Standard (not cost-optimized), using only Deep Archive (retrieval time issue), and EBS gp3 (not an S3 storage class) all conflict with the requirements and are incorrect. The distinction with Lifecycle, minimum retention periods, and the retrieval times and pricing of each tier are also asked in higher-level exams.