KodeKloud Notes

Amazon S3 offers multiple storage classes tailored to different access patterns, durability needs, and cost objectives. Selecting the right class helps you optimize costs and ensure your data is available when you need it.

Overview of S3 Storage Classes

Storage Class	Access Pattern	Durability	Availability Zones	Retrieval	Minimum Duration	Use Case
S3 Standard	Frequent	11 nines (99.999999999%)	≥3 AZs	Milliseconds	None	Website hosting, hot data
S3 Standard-IA	Infrequent	11 nines	≥3 AZs	Milliseconds	30 days	Long-lived but infrequent data
S3 One Zone-IA	Infrequent	11 nines (in one AZ)	1 AZ	Milliseconds	30 days	Secondary backups, infrequent
S3 Glacier Instant Retrieval	Rare, immediate	11 nines	≥3 AZs	Milliseconds	90 days	Archive with immediate access
S3 Glacier Flexible Retrieval	Rare, delayed	11 nines	≥3 AZs	Minutes–Hours	90 days	Cost-optimized archives
S3 Glacier Deep Archive	Almost never	11 nines	≥3 AZs	Hours–Days	180 days	Compliance, long-term retention
S3 Intelligent-Tiering	Variable	11 nines	≥3 AZs	Automatic	None	Automated cost optimization

S3 Standard

The default class for frequently accessed (“hot”) data:

Durability: 99.999999999%
Availability: Tolerates two concurrent AZ failures
Latency: Millisecond reads/writes
Pricing:
- Storage: per GB/month
- Data ingress: Free
- Data egress: per GB (no retrieval fee)

The image illustrates the AWS S3 Standard storage model, showing data replication across three availability zones in the "us-east-1" region, with charges based on gigabytes used.

S3 Standard-IA (Infrequent Access)

For long-lived data accessed less frequently:

Same durability as S3 Standard
Millisecond access latency
Pricing:
- Storage: lower per GB than Standard
- Retrieval fee: per GB + egress rate
- Minimum duration: 30 days
- Minimum object size: 128 KB

Standard-IA is perfect for backups and replicas that you need fast access to but rarely retrieve.

S3 One Zone-IA

Cost-optimized for infrequently accessed data not requiring multi-AZ resilience:

Durability: Stored in a single AZ
Latency: Millisecond access
Pricing:
- Storage: lowest among IA classes
- Retrieval fee: per GB + egress rate
- Minimum duration: 30 days
- Minimum object size: 128 KB

Warning

One Zone-IA stores data in a single Availability Zone. If that AZ fails, your objects become unavailable.

The image is an infographic about AWS S3 One Zone-IA, highlighting its design for infrequently accessed data, charging per GB, retrieval fees, and minimum charges, with a focus on a single availability zone.

S3 Glacier Instant Retrieval

Archival storage with immediate access for rare use cases:

Durability & resiliency: Same as Standard (≥3 AZs)
Latency: Millisecond retrieval
Pricing:
- Storage: archival rate per GB
- Retrieval fee: per GB
- Minimum duration: 90 days
- Minimum object size: 128 KB

Note

Not recommended for direct website hosting—use S3 Standard for public assets.

The image is an informational graphic about AWS S3 Glacier-Instant, highlighting it as a low-cost storage option for rarely accessed data with details on charges and availability zones.

S3 Glacier Flexible Retrieval

Previously known as “Glacier,” suitable for archive data with retrieval delays:

Availability: Objects must be restored before access
Pricing:
- Storage: slightly lower per GB than Standard-IA
- Retrieval fee: varies by retrieval tier
- Minimum duration: 90 days
- Minimum object size: 40 KB

Retrieval Tiers

Expedited: 1–5 minutes
Standard: 3–5 hours
Bulk: 5–12 hours

Restored objects are copied to Standard-IA for the duration of your restore request.

S3 Glacier Deep Archive

The lowest-cost class for data accessed once or twice a year:

Availability: Retrieval can take hours or days
Pricing:
- Storage: lowest in Amazon S3
- Retrieval fee: per GB + tier fee
- Minimum duration: 180 days
- Minimum object size: 40 KB

Retrieval Tiers

Standard: up to 12 hours
Bulk: up to 48 hours

The image is an infographic about AWS S3 Glacier Deep Archive, highlighting its options, retrieval times, charges, and storage features across availability zones. It emphasizes being the cheapest storage class in S3 with specific retrieval fees and minimum charges.

S3 Intelligent Tiering

Automatically optimizes costs for data with unknown or changing access patterns:

Automated monitoring moves objects between frequent & infrequent tiers
No retrieval fees for tier transitions
Monitoring & automation charge: per 1,000 objects

Note

Enable Intelligent Tiering to reduce manual effort and optimize monthly storage bills.

The image describes S3 Intelligent Tiering, highlighting its ability to reduce storage costs by moving data to cost-effective tiers and mentioning additional monitoring costs per 1,000 objects.

Selecting the Right S3 Storage Class

Use the decision flowchart below to pick the optimal storage class based on access frequency and immediacy:

Immediate (millisecond) access?
- Yes → Frequency?
  - Frequent → S3 Standard
  - Infrequent → Multi-AZ?
    - Yes → Standard-IA
    - No → One Zone-IA
  - Rare → Glacier Instant Retrieval
- No → Frequency?
  - Almost never → Glacier Deep Archive
  - Occasionally → Glacier Flexible Retrieval

The image is a flowchart for selecting storage options based on access frequency and immediacy, featuring options like Standard, Glacier Instant, and Glacier Deep Archive.

Additional Resources

Watch Video

Watch video content