S3 Storage Classes

Amazon S3 offers a range of storage classes designed for different access patterns and cost requirements. Choosing the right storage class is critical for optimizing costs while meeting performance needs.

Storage Class Overview

Key Principle

S3 storage classes trade off access speed and cost. Frequently accessed data should use Standard; rarely accessed data should use cheaper archive tiers.

Frequently Accessed Data

S3 Standard

The default storage class for frequently accessed data.

Attribute	Value
Durability	99.999999999% (11 9's)
Availability	99.99%
Availability Zones	≥3
Minimum Storage Duration	None
Minimum Object Size	None
Retrieval Fee	None
First Byte Latency	Milliseconds

Use Cases:

Active application data
Dynamic website content
Mobile and gaming applications
Big data analytics
Content distribution

Upload to Standard (default)

aws s3 cp file.txt s3://my-bucket/

S3 Intelligent-Tiering

Automatically moves objects between tiers based on access patterns.

Tier	Access Pattern	Storage Cost
Frequent Access	Accessed recently	Same as Standard
Infrequent Access	Not accessed for 30 days	40% lower
Archive Instant Access	Not accessed for 90 days	68% lower
Archive Access	Not accessed for 90+ days (optional)	71% lower
Deep Archive Access	Not accessed for 180+ days (optional)	95% lower

There's a small monthly monitoring and automation charge per object. Best for objects >128KB.

Upload with Intelligent-Tiering

aws s3 cp file.txt s3://my-bucket/ \
  --storage-class INTELLIGENT_TIERING

Configure archive access tiers:

Enable archive access tiers

aws s3api put-bucket-intelligent-tiering-configuration \
  --bucket my-bucket \
  --id my-config \
  --intelligent-tiering-configuration '{
    "Id": "my-config",
    "Status": "Enabled",
    "Tierings": [
      {
        "Days": 90,
        "AccessTier": "ARCHIVE_ACCESS"
      },
      {
        "Days": 180,
        "AccessTier": "DEEP_ARCHIVE_ACCESS"
      }
    ]
  }'

Infrequently Accessed Data

S3 Standard-Infrequent Access (Standard-IA)

For data accessed less frequently but requiring rapid access when needed.

Attribute	Value
Durability	99.999999999% (11 9's)
Availability	99.9%
Availability Zones	≥3
Minimum Storage Duration	30 days
Minimum Object Size	128KB (charged as 128KB if smaller)
Retrieval Fee	Per GB retrieved
First Byte Latency	Milliseconds

Cost Comparison (vs Standard):

Storage: ~40% cheaper
Retrieval: Additional fee per GB

Use Cases:

Backups and disaster recovery
Long-term storage that needs quick access
Older data that's occasionally accessed

Upload to Standard-IA

aws s3 cp file.txt s3://my-bucket/ \
  --storage-class STANDARD_IA

S3 One Zone-Infrequent Access (One Zone-IA)

Same as Standard-IA but stored in a single Availability Zone.

Attribute	Value
Durability	99.999999999% (11 9's) within AZ
Availability	99.5%
Availability Zones	1
Minimum Storage Duration	30 days
Minimum Object Size	128KB
Retrieval Fee	Per GB retrieved
First Byte Latency	Milliseconds

Data is lost if the Availability Zone is destroyed. Not recommended for sole copies of critical data.

Cost Comparison (vs Standard-IA):

Storage: ~20% cheaper than Standard-IA

Use Cases:

Secondary backups
Easily reproducible data
Thumbnails or transcoded media

Upload to One Zone-IA

aws s3 cp file.txt s3://my-bucket/ \
  --storage-class ONEZONE_IA

Archive Storage Classes

S3 Glacier Instant Retrieval

Archive storage with millisecond retrieval.

Attribute	Value
Durability	99.999999999% (11 9's)
Availability	99.9%
Availability Zones	≥3
Minimum Storage Duration	90 days
Minimum Object Size	128KB
Retrieval Fee	Per GB
First Byte Latency	Milliseconds

Cost Comparison (vs Standard):

Storage: ~68% cheaper
Retrieval: Higher fee than Standard-IA

Use Cases:

Medical images
News media archives
User-generated content archives
Accessed once per quarter or less

Upload to Glacier Instant Retrieval

aws s3 cp file.txt s3://my-bucket/ \
  --storage-class GLACIER_IR

S3 Glacier Flexible Retrieval (formerly Glacier)

Low-cost archive storage with flexible retrieval times.

Attribute	Value
Durability	99.999999999% (11 9's)
Availability	99.99% (after restore)
Availability Zones	≥3
Minimum Storage Duration	90 days
Minimum Object Size	40KB overhead
First Byte Latency	Minutes to hours

Retrieval Options:

Tier	Time	Use Case
Expedited	1-5 minutes	Urgent access (costs most)
Standard	3-5 hours	Regular retrieval
Bulk	5-12 hours	Large data sets (cheapest)

Upload to Glacier Flexible

aws s3 cp file.txt s3://my-bucket/ \
  --storage-class GLACIER

Restore an object:

Restore from Glacier

# Standard retrieval, available for 7 days
aws s3api restore-object \
  --bucket my-bucket \
  --key my-archive.zip \
  --restore-request '{
    "Days": 7,
    "GlacierJobParameters": {
      "Tier": "Standard"
    }
  }'

S3 Glacier Deep Archive

Lowest cost storage for long-term retention.

Attribute	Value
Durability	99.999999999% (11 9's)
Availability	99.99% (after restore)
Availability Zones	≥3
Minimum Storage Duration	180 days
Minimum Object Size	40KB overhead
First Byte Latency	12-48 hours

Retrieval Options:

Tier	Time	Use Case
Standard	Within 12 hours	Regular retrieval
Bulk	Within 48 hours	Large data sets (cheapest)

Cost Comparison:

~95% cheaper than S3 Standard
Cheapest AWS storage option

Use Cases:

Compliance archives (7+ years retention)
Digital preservation
Magnetic tape replacement
Healthcare and financial records

Upload to Glacier Deep Archive

aws s3 cp file.txt s3://my-bucket/ \
  --storage-class DEEP_ARCHIVE

Restore from Deep Archive

aws s3api restore-object \
  --bucket my-bucket \
  --key compliance-record.pdf \
  --restore-request '{
    "Days": 30,
    "GlacierJobParameters": {
      "Tier": "Bulk"
    }
  }'

S3 Express One Zone

Newest Storage Class

S3 Express One Zone is designed for performance-critical applications requiring single-digit millisecond latency.

Attribute	Value
Latency	Single-digit milliseconds
Requests/sec	Hundreds of thousands
Availability Zones	1 (choose your AZ)
Storage Type	Directory buckets

Use Cases:

Machine learning training
Financial modeling
Real-time analytics
High-performance computing

Create Express One Zone directory bucket

aws s3api create-bucket \
  --bucket my-bucket--usw2-az1--x-s3 \
  --region us-west-2 \
  --create-bucket-configuration '{
    "Location": {
      "Type": "AvailabilityZone",
      "Name": "usw2-az1"
    },
    "Bucket": {
      "DataRedundancy": "SingleAvailabilityZone",
      "Type": "Directory"
    }
  }'

Reduced Redundancy (Deprecated)

Not Recommended

S3 Reduced Redundancy Storage (RRS) is deprecated. Use S3 One Zone-IA or S3 Standard instead.

Storage Class Comparison Table

Class	Durability	Availability	AZs	Min Duration	Min Size	Latency
Standard	11 9's	99.99%	≥3	None	None	ms
Intelligent-Tiering	11 9's	99.9%	≥3	None	None	ms
Standard-IA	11 9's	99.9%	≥3	30 days	128KB	ms
One Zone-IA	11 9's	99.5%	1	30 days	128KB	ms
Glacier Instant	11 9's	99.9%	≥3	90 days	128KB	ms
Glacier Flexible	11 9's	99.99%*	≥3	90 days	40KB+	min-hrs
Glacier Deep Archive	11 9's	99.99%*	≥3	180 days	40KB+	hrs
Express One Zone	11 9's	99.95%	1	1 hour	None	< 10ms

*After restore

Changing Storage Classes

Copy object to new storage class

aws s3 cp s3://my-bucket/file.txt s3://my-bucket/file.txt \
  --storage-class GLACIER_IR

Using S3 API

aws s3api copy-object \
  --bucket my-bucket \
  --key file.txt \
  --copy-source my-bucket/file.txt \
  --storage-class STANDARD_IA \
  --metadata-directive COPY

Use S3 Batch Operations for large-scale storage class changes:

batch-job-manifest.json

{
  "Spec": {
    "Format": "S3BatchOperations_CSV_20180820",
    "Fields": ["Bucket", "Key"]
  },
  "Location": {
    "ObjectArn": "arn:aws:s3:::my-bucket/manifest.csv",
    "ETag": "abc123"
  }
}

Create batch job

aws s3control create-job \
  --account-id 123456789012 \
  --manifest file://manifest.json \
  --operation '{
    "S3PutObjectCopy": {
      "TargetResource": "arn:aws:s3:::my-bucket",
      "StorageClass": "GLACIER"
    }
  }' \
  --report '{
    "Bucket": "arn:aws:s3:::report-bucket",
    "Enabled": true,
    "Format": "Report_CSV_20180820",
    "ReportScope": "AllTasks"
  }' \
  --role-arn arn:aws:iam::123456789012:role/S3BatchRole \
  --priority 1

Automate storage class transitions:

lifecycle-config.json

{
  "Rules": [
    {
      "ID": "OptimizeStorage",
      "Status": "Enabled",
      "Filter": {},
      "Transitions": [
        {
          "Days": 30,
          "StorageClass": "STANDARD_IA"
        },
        {
          "Days": 90,
          "StorageClass": "GLACIER_IR"
        },
        {
          "Days": 365,
          "StorageClass": "DEEP_ARCHIVE"
        }
      ]
    }
  ]
}

Apply lifecycle configuration

aws s3api put-bucket-lifecycle-configuration \
  --bucket my-bucket \
  --lifecycle-configuration file://lifecycle-config.json

Cost Optimization Strategies

Analyze Access Patterns

Use S3 Storage Lens and access logs to understand how data is accessed:

Create Storage Lens configuration

aws s3control put-storage-lens-configuration \
  --account-id 123456789012 \
  --config-id my-dashboard \
  --storage-lens-configuration '{
    "Id": "my-dashboard",
    "IsEnabled": true,
    "AccountLevel": {
      "BucketLevel": {
        "ActivityMetrics": {
          "IsEnabled": true
        }
      }
    }
  }'

Use Intelligent-Tiering for Unknown Patterns

When you can't predict access patterns:

Set default storage class

# In lifecycle rule
{
  "Rules": [{
    "ID": "DefaultToIntelligentTiering",
    "Filter": {"Prefix": ""},
    "Status": "Enabled",
    "Transitions": [{
      "Days": 0,
      "StorageClass": "INTELLIGENT_TIERING"
    }]
  }]
}

Set Lifecycle Rules Based on Data Age

Create rules matching your retention requirements:

Tiered lifecycle example

{
  "Rules": [
    {
      "ID": "LogsLifecycle",
      "Status": "Enabled",
      "Filter": {"Prefix": "logs/"},
      "Transitions": [
        {"Days": 30, "StorageClass": "STANDARD_IA"},
        {"Days": 90, "StorageClass": "GLACIER"}
      ],
      "Expiration": {"Days": 365}
    }
  ]
}

Monitor and Adjust

Regularly review storage costs and adjust policies:

Get bucket storage metrics

aws cloudwatch get-metric-statistics \
  --namespace AWS/S3 \
  --metric-name BucketSizeBytes \
  --dimensions Name=BucketName,Value=my-bucket Name=StorageType,Value=StandardStorage \
  --start-time 2024-01-01T00:00:00Z \
  --end-time 2024-01-31T23:59:59Z \
  --period 86400 \
  --statistics Average

Best Practices

Storage Class Selection Guidelines

Standard: Default for active data with unpredictable access
Intelligent-Tiering: Best for large objects with unknown access patterns
Standard-IA: Data accessed monthly, needs quick access
One Zone-IA: Reproducible data, secondary copies
Glacier Instant: Quarterly access, millisecond retrieval needed
Glacier Flexible: Annual access, can wait hours
Deep Archive: Rarely accessed, compliance/legal requirements

S3 Storage Classes

Storage Class Overview

Frequently Accessed Data

S3 Standard

S3 Intelligent-Tiering

Infrequently Accessed Data

S3 Standard-Infrequent Access (Standard-IA)

S3 One Zone-Infrequent Access (One Zone-IA)

Archive Storage Classes

S3 Glacier Instant Retrieval

S3 Glacier Flexible Retrieval (formerly Glacier)

S3 Glacier Deep Archive

S3 Express One Zone

Reduced Redundancy (Deprecated)

Storage Class Comparison Table

Changing Storage Classes

Cost Optimization Strategies

Analyze Access Patterns

Use Intelligent-Tiering for Unknown Patterns

Set Lifecycle Rules Based on Data Age

Monitor and Adjust

Best Practices

Next Steps

Buckets

Lifecycle Rules

Security

S3 CLI Reference

On this page