Storage Classes

NFYio offers multiple storage classes to balance cost and access patterns. Choose the right class for each object or use lifecycle rules to automatically move data to cheaper tiers as it ages.

Available Storage Classes

Class	Use Case	Access Latency	Retrieval	Cost
STANDARD	Frequently accessed data	Low	Immediate	Highest
STANDARD_IA	Infrequent Access — backups, archives	Low	Immediate	Lower
GLACIER	Long-term archival, rarely accessed	Higher	Minutes to hours	Lowest

Standard

• Best for: Active data, hot storage, frequently read/written objects
• Access: Immediate, low latency
• Use when: Objects are accessed regularly (e.g., user uploads, application assets)

Infrequent Access (STANDARD_IA)

• Best for: Backups, logs, data accessed less than once a month
• Access: Immediate, same as Standard
• Cost: Lower storage cost; may have retrieval fees
• Use when: You need occasional access but want to reduce storage costs

Glacier

• Best for: Long-term archival, compliance, data rarely or never retrieved
• Access: Retrieval can take minutes to hours (restore process)
• Cost: Lowest storage cost; retrieval has higher cost and delay
• Use when: Data is stored for compliance or disaster recovery, not regular access

Setting Storage Class

On Upload

AWS CLI:

aws s3 cp backup.tar.gz s3://my-bucket/archives/backup.tar.gz \
  --storage-class STANDARD_IA \
  --endpoint-url https://storage.yourdomain.com

JavaScript:

await client.send(new PutObjectCommand({
  Bucket: 'my-bucket',
  Key: 'archives/backup.tar.gz',
  Body: fileStream,
  StorageClass: 'STANDARD_IA',
}));

Default Bucket Storage Class

Set a default storage class for the bucket so new objects use it unless overridden:

aws s3api put-bucket-storage-class \
  --bucket my-bucket \
  --storage-class STANDARD_IA \
  --endpoint-url https://storage.yourdomain.com

Lifecycle Transitions

Automatically move objects between storage classes based on age. Define rules in the bucket lifecycle configuration.

Example: Logs Bucket

Move logs to Infrequent Access after 30 days, then to Glacier after 90 days:

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

Example: Backup Retention

Keep backups in Standard for 7 days (recent restores), then move to Glacier for 1 year:

{
  "Rules": [
    {
      "ID": "BackupLifecycle",
      "Status": "Enabled",
      "Filter": { "Prefix": "backups/" },
      "Transitions": [
        { "Days": 7, "StorageClass": "STANDARD_IA" },
        { "Days": 30, "StorageClass": "GLACIER" }
      ],
      "Expiration": { "Days": 365 }
    }
  ]
}

Versioned Buckets

Transition noncurrent versions (old versions after overwrite or delete):

{
  "Rules": [
    {
      "ID": "OldVersionsToGlacier",
      "Status": "Enabled",
      "Filter": {},
      "NoncurrentVersionTransitions": [
        { "NoncurrentDays": 30, "StorageClass": "STANDARD_IA" },
        { "NoncurrentDays": 90, "StorageClass": "GLACIER" }
      ],
      "NoncurrentVersionExpiration": { "NoncurrentDays": 365 }
    }
  ]
}

Applying Lifecycle Configuration

aws s3api put-bucket-lifecycle-configuration \
  --bucket my-bucket \
  --lifecycle-configuration file://lifecycle.json \
  --endpoint-url https://storage.yourdomain.com

Glacier Retrieval

Objects in Glacier require a restore before download. Initiate a restore request; when ready, the object is temporarily available for download.

Restore request (if supported by your deployment):

aws s3api restore-object \
  --bucket my-bucket \
  --key archives/old-backup.tar.gz \
  --restore-request '{"Days": 7}' \
  --endpoint-url https://storage.yourdomain.com

After the restore completes (check object metadata or wait for the configured days), download normally. The restored copy may expire after the specified period.

Cost Optimization Tips

1. Use lifecycle rules — Automate transitions; don’t manually move objects
2. Prefix-based rules — Apply different policies to different paths (e.g., uploads/ vs archives/)
3. Review access patterns — If data is rarely read, move to STANDARD_IA or GLACIER sooner
4. Versioning + lifecycle — Transition old versions to cheaper classes; expire delete markers
5. Delete what you don’t need — Use expiration rules to remove temporary or obsolete data

Next Steps

• Managing Buckets — Configure lifecycle rules
• Object Versioning — Version lifecycle and noncurrent transitions
• Working with Objects — Upload with storage class