SAA-C03 Practice Question 80: S3 One Zone IA Derived Data

One Zone-IA stores data in a single Availability Zone, which makes it cheaper than Standard-IA, and it still serves objects with millisecond access. Its one drawback is that the data is not resilient to the loss of that Availability Zone, and AWS's own recommendation is to use it exactly when you can re-create the data if the Availability Zone fails. Thumbnails regenerated from an original in a couple of seconds are that case. Transitioning after 30 days also clears the class's 30-day minimum storage duration and matches the point at which access drops off.

Option A saves the most on storage but changes the access behaviour. Every later request would have to regenerate the image, which takes seconds rather than milliseconds, failing the stated latency requirement, and it adds compute cost on every miss.

Option B keeps millisecond retrieval but is the wrong class for this access pattern. Glacier Instant Retrieval is priced for archives touched about once a quarter and charges a per-GB retrieval fee on top, along with a 90-day minimum duration.

Option C is what you choose when the access pattern is unknown or changing. Here it is known (heavy for a month, then negligible), so paying the per-object monitoring and automation fee buys a decision the team has already made, and Intelligent-Tiering's tiers are all multi-Availability-Zone, so it never reaches the cheaper single-zone price.

Reference: Understanding and managing Amazon S3 storage classes