Why Storage Classes Exist
Not all data is accessed the same way: a product image served on every page load has completely different economics than a compliance log you must keep for seven years but will almost never open. S3 storage classes let you pay for exactly the durability, availability, and retrieval speed each object needs, all while keeping the same API, the same bucket, and the same key — only the storage class metadata on the object changes. Every class still gives you 11 nines of durability (except One Zone-IA, which is single-AZ); what differs is availability SLA, minimum storage duration, retrieval latency, and per-GB versus per-request pricing.
Cricket analogy: It's like a franchise keeping this season's playing-XI kit ready in the dressing room (hot storage) versus retired legends' jerseys boxed in a stadium archive (cold storage) — both are the same 'kit inventory system,' just accessed at very different frequencies.
The Standard, Infrequent Access, and One Zone Tiers
S3 Standard is for frequently accessed, latency-sensitive data with no retrieval fee and millisecond first-byte latency. S3 Standard-IA (Infrequent Access) and S3 One Zone-IA cost roughly 40-46% less per GB stored but charge a per-GB retrieval fee and enforce a 30-day minimum storage duration, so they make sense for backups or disaster-recovery copies you access maybe once a month. One Zone-IA goes further by storing data in a single Availability Zone rather than three, cutting cost further but meaning an AZ-level disaster can destroy the data — appropriate only for easily-recreatable data like thumbnails re-derivable from a Standard-tier original.
Cricket analogy: It's like a team keeping its main training kit at the primary academy (Standard) versus a satellite backup kit stored at just one regional center (One Zone-IA) — cheaper, but if that one center floods, the backup kit is gone.
Glacier and Intelligent-Tiering
For true archival data, S3 Glacier Instant Retrieval offers millisecond access at Glacier-level pricing for data touched about once a quarter; S3 Glacier Flexible Retrieval offers retrieval in minutes to hours (Expedited, Standard, Bulk speed tiers) for data accessed once or twice a year; and S3 Glacier Deep Archive is the cheapest tier available, with 12-hour standard restore times, aimed at 7-10 year compliance retention. S3 Intelligent-Tiering removes the guesswork: it monitors access patterns per object and automatically moves objects between frequent, infrequent, and archive tiers for a small monitoring fee, which is ideal when access patterns are unknown or unpredictable.
Cricket analogy: It's like the ICC keeping World Cup final footage instantly streamable (Glacier Instant Retrieval) versus 1990s domestic match tapes needing hours to pull from a physical vault (Glacier Flexible Retrieval) versus 1970s footage needing a full day's notice (Deep Archive).
{
"Rules": [
{
"ID": "MoveOldLogsToArchive",
"Filter": { "Prefix": "logs/" },
"Status": "Enabled",
"Transitions": [
{ "Days": 30, "StorageClass": "STANDARD_IA" },
{ "Days": 90, "StorageClass": "GLACIER" },
{ "Days": 365, "StorageClass": "DEEP_ARCHIVE" }
],
"Expiration": { "Days": 2555 }
}
]
}Lifecycle rules like the one above are the standard way to automate storage-class transitions: apply the JSON policy with 'aws s3api put-bucket-lifecycle-configuration'. Objects smaller than 128 KB are not worth transitioning to IA/Glacier tiers because the per-object overhead can exceed the storage savings.
Standard-IA, One Zone-IA, and Glacier tiers charge an early-deletion fee if you delete or transition an object before its minimum storage duration (30 days for IA tiers, 90 for Glacier Flexible Retrieval, 180 for Deep Archive). Frequently-changing or short-lived objects should stay on S3 Standard.
- Storage classes trade retrieval latency and per-request fees against per-GB storage cost, all under the same S3 API.
- S3 Standard: no retrieval fee, millisecond latency, multi-AZ — for frequently accessed data.
- Standard-IA and One Zone-IA: cheaper storage, per-GB retrieval fee, 30-day minimum duration; One Zone-IA is single-AZ.
- Glacier Instant Retrieval: archive pricing with millisecond access, for data touched roughly quarterly.
- Glacier Flexible Retrieval and Deep Archive: cheapest tiers, minutes-to-hours or up to 12-hour restore times, for compliance archives.
- S3 Intelligent-Tiering automatically moves objects between tiers based on observed access patterns for a small monitoring fee.
- Lifecycle policies automate transitions and expirations; watch for early-deletion fees on IA and Glacier tiers.
Practice what you learned
1. Which S3 storage class stores data in only a single Availability Zone?
2. What is the minimum storage duration before which an early-deletion fee applies for S3 Standard-IA?
3. Which storage class is best suited for data whose access pattern is unknown or unpredictable?
4. Which Glacier tier offers millisecond retrieval latency at archive-level storage pricing?
5. What AWS mechanism automates moving objects between storage classes over time?
Was this page helpful?
You May Also Like
S3 Fundamentals
Learn how Amazon S3 stores data as objects in buckets, and the durability, consistency, and access-control model that makes it the backbone of AWS storage.
AWS Storage Gateway Overview
See how AWS Storage Gateway bridges on-premises applications to S3, EBS snapshots, and Glacier using File, Volume, and Tape gateway modes.
EFS Explained
Learn how Amazon EFS provides a fully managed, elastic NFS file system that multiple EC2 instances and containers can mount and share concurrently.