A county clerk must archive closed case PDFs that are rarely opened but, when needed, must appear within milliseconds for courtroom staff. Which Amazon S3 storage class best balances archive cost with that restore latency?

S3 Glacier Instant Retrieval is the correct storage class for closed case PDFs that are rarely opened but must appear within milliseconds for courtroom staff because it targets long-lived, infrequently accessed objects that still require millisecond GET latency. The county clerk can archive at a lower storage price than hot classes while meeting the immediacy constraint of courtroom pulls. Glacier Deep Archive for every object because it is always the lowest price fails conceptually: Deep Archive minimizes storage cost but restore times are measured in hours, which breaks the millisecond courtroom requirement even if unit price looks attractive. S3 Standard-Infrequent Access with no lifecycle because archives never change tiers fails because Standard-IA is not the purpose-built archive class for this rare-access-plus-instant-retrieve pattern, and refusing lifecycle review ignores ongoing cost optimization as objects age. S3 One Zone-Infrequent Access in a single Availability Zone for courtroom HA fails because One Zone-IA stores data in a single AZ, reducing resilience relative to multi-AZ designs inappropriate for critical courtroom retrievals sold as high availability. Exam caveat: when the stem demands archive economics and millisecond retrieval, choose Glacier Instant Retrieval over Deep Archive, generic IA without lifecycle thinking, or single-AZ IA. Operational check: classify closed-case PDFs by access frequency and required latency; apply Glacier Instant Retrieval (or a lifecycle that lands there); sample restore/GET timing with courtroom stakeholders; and document that Deep Archive remains reserved for true offline retention without immediacy needs.