Skip to content

Storage Cost Optimization — S3, EBS, Snapshot, Backup#

"S3 bill $20K/mo, 80% of it is 6-month-old logs + zombie snapshots. Lifecycle policy + tier transition = same data 70% cheaper. 4 hours of work, a permanent $14K/mo savings."

This guide covers cost optimization techniques for S3 / EBS / snapshots / backups, with concrete commands and lifecycle policies.

🎯 Storage Cost Drivers#

Cost source % common
EBS volumes (idle) 20-40%
EBS snapshots (old, unused) 15-30%
S3 Standard (old logs/files) 20-40%
NAT Gateway egress 10-25%
Backup retention (uncontrolled) 15-30%

🪣 1️⃣ S3 Lifecycle Policy#

Staged tier#

LifecycleRules:
  - Filter: {Prefix: logs/}
    Transitions:
      - Days: 30
        StorageClass: STANDARD_IA      # 40% cheaper
      - Days: 90
        StorageClass: GLACIER          # 80% cheaper
      - Days: 365
        StorageClass: DEEP_ARCHIVE     # 95% cheaper
    Expiration: {Days: 2555}            # 7 years

Cost comparison (per TB per month)#

STANDARD:        $23
STANDARD_IA:     $12.5
GLACIER:         $4
GLACIER_IR:      $5      (instant retrieval)
DEEP_ARCHIVE:    $1

Use case mapping#

Data type Recommended class
Active data STANDARD
Old logs (30+ days, occasional access) STANDARD_IA
Compliance audit (1+ year) GLACIER or DEEP_ARCHIVE
Backup (1+ year) GLACIER + Object Lock

Intelligent-Tiering#

StorageClass: INTELLIGENT_TIERING

→ AWS shifts tiers automatically (tier transition fee $0.0025/1K objects).

💾 2️⃣ EBS Optimization#

Idle volume detection#

# Detached volumes
aws ec2 describe-volumes \
  --filters "Name=status,Values=available" \
  --query 'Volumes[*].[VolumeId, Size, CreateTime]' \
  --output table

# Idle but attached (CloudWatch read/write 0)
aws cloudwatch get-metric-statistics \
  --namespace AWS/EBS \
  --metric-name VolumeReadOps \
  --dimensions Name=VolumeId,Value=vol-xxx \
  --start-time $(date -d '14 days ago' +%FT%T) \
  --end-time $(date +%FT%T) \
  --period 86400 \
  --statistics Sum

gp3 migration (from gp2)#

# gp2 → gp3: 20% cheaper + better IOPS
aws ec2 modify-volume \
  --volume-id vol-xxx \
  --volume-type gp3 \
  --iops 3000 --throughput 125

🔑 gp3 default is cheaper than gp2. Migration is easy, instant.

Right-sizing#

  • 100 GB volume, 30% usage → shrink to 50 GB (tooling: shrink lvm/xfs/ext4)

  • 80% usage → grow (gp3 instant)

📸 3️⃣ Snapshot Cleanup#

Old snapshots#

# 90+ days old
aws ec2 describe-snapshots --owner-ids self \
  --query 'Snapshots[?StartTime<=`2026-02-04`].[SnapshotId, StartTime, Description]' \
  --output table

# Automatic delete (Cloud Custodian policy)

# custodian-snapshot-cleanup.yml
policies:
  - name: ebs-snapshot-old
    resource: ebs-snapshot
    filters:
      - type: age
        days: 90
        op: gt
      - "tag:DoNotDelete": absent
    actions:
      - delete

custodian run -s out custodian-snapshot-cleanup.yml

Lifecycle Manager (DLM)#

# AWS DLM: scheduled snapshot
ResourceTypes: [VOLUME]
Schedules:
  - Name: daily
    CreateRule: {Interval: 24, IntervalUnit: HOURS}
    RetainRule: {Count: 7}    # keep for 7 days

🌐 4️⃣ Egress Cost (NAT Gateway / Inter-AZ)#

Egress cost#

S3 GET (within AWS region):    $0
S3 GET (cross region):         $0.02/GB
EC2 → Internet:                $0.09/GB
NAT Gateway:                   $0.045/GB
Inter-AZ:                      $0.01/GB

Optimization#

  1. VPC Endpoints (NAT bypass for S3, DynamoDB) 

    aws ec2 create-vpc-endpoint \
  --vpc-id vpc-xxx \
  --service-name com.amazonaws.<REGION>.s3 \
  --route-table-ids rtb-xxx
    → S3 traffic bypasses the NAT, $0.

  2. Minimize cross-AZ traffic: pod anti-affinity, single-AZ for non-HA

  3. CDN: edge cache → origin traffic drops 50%+

📦 5️⃣ Backup Retention Discipline#

Rule#

  • Production daily backup: 30 days
  • Weekly: 12 weeks
  • Monthly: 12 months
  • Yearly (compliance): 7 years

Automatic cleanup#

# AWS Backup
BackupPlanRule:
  RuleName: daily
  TargetBackupVaultName: prod-vault
  ScheduleExpression: cron(0 2 * * ? *)
  Lifecycle:
    DeleteAfterDays: 30
    MoveToColdStorageAfterDays: 7   # cheap tier

  RuleName: monthly
  ScheduleExpression: cron(0 2 1 * ? *)
  Lifecycle:
    DeleteAfterDays: 365
    MoveToColdStorageAfterDays: 30

🔍 Storage Cost Audit#

Quarterly review#

# Top 10 cost-heavy bucket
aws s3 ls --summarize --human-readable

# Is there a lifecycle policy?
aws s3api get-bucket-lifecycle-configuration --bucket <BUCKET>

# Versioning storage cost
aws s3api list-object-versions --bucket <BUCKET>

Cost Explorer#

aws ce get-cost-and-usage \
  --time-period Start=2026-04-01,End=2026-05-01 \
  --granularity MONTHLY \
  --metrics BlendedCost \
  --group-by Type=DIMENSION,Key=USAGE_TYPE \
  --filter '{"Dimensions": {"Key": "SERVICE", "Values": ["Amazon S3"]}}'

🚫 Anti-Pattern Table#

Anti-pattern Why it's bad Do this instead
No lifecycle policy 70% overpayment 30/90/365-day tier
Snapshot retention unbounded Piles up DLM + retention
gp2 still in use 20% overpayment Migrate to gp3
Egress heavy through NAT $$ VPC Endpoints
Cross-AZ pod placement random Inter-AZ traffic Anti-affinity strategy
No backup retention / unbounded Storage cost + KVKK (Turkey's Personal Data Protection Law, No. 6698) violation Disciplined retention
No S3 versioning + lifecycle Versions pile up Lifecycle for versioned objects
No Custodian / cleanup automation Manual neglect Cron + script
Detached volume for months Idle cost Cleanup policy
Intelligent Tiering treated as mandatory Tier fee isn't always minimal Classic lifecycle is cheaper in some cases

📋 Storage Cost Optimization Checklist#

[ ] S3 lifecycle: 30/90/365-day tier
[ ] S3 Intelligent Tiering (large buckets)
[ ] S3 versioning lifecycle (old version cleanup)
[ ] EBS gp3 migration (from gp2)
[ ] EBS detached volume cleanup (Custodian)
[ ] EBS snapshot retention (DLM, 7-30 days)
[ ] Snapshot cleanup (90+ days old)
[ ] VPC Endpoints (S3, DynamoDB)
[ ] NAT Gateway minimize
[ ] CDN: static assets to the edge
[ ] Backup retention disciplined (3-2-1 + lifecycle)
[ ] Quarterly storage cost review
[ ] Per-bucket cost dashboard
[ ] Tag policy: cost-center per bucket

📚 References#

"Storage cost is assumed to be 'small' — 6 months later it's 30% of the bill. Lifecycle policy + cleanup automation = free savings, one week of implementation."