Amazon S3 Exam frequency ⭐⭐⭐⭐⭐

S3 Storage Classes & Lifecycle

Amazon S3 Storage Classes and Lifecycle Management

S3Storage ClassesLifecycleSAA-C03
Last organized

Choose an S3 storage class from access frequency, retrieval time, availability scope, and total cost, then automate transitions and expiration with Lifecycle.

Basic Information

FieldDetails
English nameS3 Storage Classes & Lifecycle
Full nameAmazon S3 Storage Classes and Lifecycle Management
Chinese descriptionS3 存储类别与生命周期
Japanese descriptionS3 ストレージクラスとライフサイクル
Exam frequency⭐⭐⭐⭐⭐
Often confused withStandard / IA / Glacier / Intelligent-Tiering

Selection logic

Ask four questions first:

  1. How often is the data accessed?
  2. Must retrieval be immediate, or can it wait?
  3. Can the data be rebuilt, and is a single-AZ risk acceptable?
  4. Beyond storage, what are the retrieval, request, and minimum-duration costs?

Main storage classes

Storage classBest fitKey idea
S3 StandardFrequently accessed data, dynamic websites, data lakesMulti-AZ, low latency, default class
S3 Intelligent-TieringUnknown or changing access patternsAutomatically moves data between access tiers; monitoring and automation fees apply
S3 Standard-IAInfrequent access with millisecond retrievalMulti-AZ, lower storage price, retrieval fees
S3 One Zone-IARe-creatable, infrequently accessed dataSingle AZ and lower cost
S3 Express One ZoneVery high performance and latency-sensitive workloadsSingle AZ, directory bucket, single-digit millisecond access
Glacier Instant RetrievalRarely accessed archives that still need millisecond retrievalImmediate access to archive data
Glacier Flexible RetrievalRetrieval can wait minutes to hoursFlexible archive retrieval
Glacier Deep ArchiveLong-term retention and very rare restoresAmong the lowest-cost tiers, with the longest restore time
S3 on OutpostsOn-premises object storageData residency, low latency, hybrid cloud

Lifecycle Policy

Rules can automatically:

  • Transition objects or old versions to lower-cost classes.
  • Expire objects after a retention period.
  • Clean up old versions and delete markers.
  • Abort incomplete multipart uploads.

Lifecycle works well when access patterns are predictable, such as “move to IA after 30 days, archive after 180 days, and delete after seven years.”

Intelligent-Tiering vs Lifecycle

  • Predictable access pattern: Lifecycle.
  • Unknown or changing access pattern: Intelligent-Tiering.
  • They can be combined, but monitoring fees, retrieval fees, minimum storage duration, and object-size limits still matter.

Common scenarios

  • Frequently used objects → Standard.
  • Infrequent but immediate retrieval → Standard-IA.
  • Rare access with millisecond retrieval → Glacier Instant Retrieval.
  • Minutes-to-hours retrieval is acceptable → Glacier Flexible Retrieval.
  • Long-term compliance retention → Deep Archive.
  • Re-creatable data with acceptable single-AZ risk → One Zone-IA.
  • Single-AZ, very high performance → Express One Zone.

Common misconceptions

  • The lowest storage price does not always produce the lowest total cost.
  • Glacier Instant Retrieval is not a slow-restore tier.
  • One Zone classes are not inherently “unsafe”; their failure scope differs from multi-AZ classes.
  • Lifecycle does not understand business value, so a bad rule can archive or delete data too early.
  • S3 on Outposts stores object data on Outposts, but compliance still depends on configuration, data flow, and governance.

Memory aid

Standard for frequent access; IA for infrequent immediate access; Glacier Instant for rare immediate access; Flexible when waiting is acceptable; Deep Archive for long retention; Intelligent-Tiering when the pattern is unknown.