All cheat sheets
Storagecheat sheet

S3 Glacier: Choosing the Right Tier

The three Glacier tiers — Instant, Flexible, Deep Archive — differ 4× in price and a million× in retrieval speed. A decision guide with the price ladder, the minimums, and the access-pattern test that picks the right one.

Last reviewed: July 11, 2026

Prefer step by step? Take the guided S3 Glacier (Flexible Retrieval) walkthrough — free, anonymous, interactive.Have an AI assistant? Jump to the copy-paste audit prompt at the bottom of this sheet.

TL;DR: "Glacier" is three different products wearing one name. Instant Retrieval reads in milliseconds at ~$0.004/GB-month; Flexible Retrieval reads in minutes-to-hours at ~$0.0036; Deep Archive reads in 12–48 hours at ~$0.00099. Pick by access pattern, not price tag — the expensive mistake is data one rung away from where it belongs, in either direction.

The numbers

Tier $/GB-month Retrieval Min. duration Min. object
Standard (reference) $0.023 ms, free GETs — —
Glacier Instant Retrieval $0.004 milliseconds, ~$0.03/GB fee 90 days 128 KB
Glacier Flexible Retrieval $0.0036 1 min–12 h by option 90 days 40 KB
Glacier Deep Archive $0.00099 12–48 h 180 days 40 KB

All three: 11-nines durability across 3+ AZs — identical to Standard. Cheap is not fragile; the difference is purely read latency and fees.

Do this

  1. Apply the access-pattern test per dataset:

    • Touched a few times a year, needed instantly (legal pulls, medical imaging, licensing) → Instant Retrieval
    • Touched rarely, hours of notice available (audits, re-edits, backup restores) → Flexible Retrieval
    • Realistically never read again (7-year retention, decommissioned systems) → Deep Archive

  2. Build the ladder as one lifecycle policy so data descends automatically:

    aws s3api put-bucket-lifecycle-configuration --bucket YOUR-BUCKET \
  --lifecycle-configuration '{"Rules":[{"ID":"ladder","Status":"Enabled",
  "Filter":{},"Transitions":[
    {"Days":30,"StorageClass":"GLACIER_IR"},
    {"Days":365,"StorageClass":"DEEP_ARCHIVE"}]}]}'

  3. Check for one-rung-early ladders: policies written years ago often stop at Glacier Flexible; multi-year retention that never restores should continue to Deep Archive (4× cheaper).

  4. Use S3 Select on restores from Flexible when you need specific records out of large archives — pulling only matching rows instead of whole objects slashes retrieval fees on compliance searches.

Gotchas

  • Wrong-tier-by-one costs real money in both directions: Deep Archive data restored monthly bleeds retrieval fees and hours; Instant Retrieval data never read overpays ~4× vs Deep Archive.
  • The minimums stack: a small object deleted early is billed at the minimum size for the minimum duration, whichever tier.
  • Transitions run once daily — plan compliance cutover dates around the lag.
  • Serving live traffic from any Glacier tier (even Instant, because of per-GB fees) is a bill you'll notice — product assets belong in Standard or Intelligent-Tiering.

Skip this if

Run this audit with your AI assistant

Paste this into Claude, ChatGPT, or any agent that can run the AWS CLI with read-only credentials. It audits your account for exactly the waste this sheet describes — and changes nothing.

You are auditing an AWS account's archival-storage tiering across the
three S3 Glacier classes. Use the AWS CLI with READ-ONLY credentials.
Do not create, modify, or delete anything — report findings and
recommended (unapplied) fixes only.

1. Inventory: per bucket, GB stored per storage class (CloudWatch
   BucketSizeBytes by StorageType: StandardStorage, StandardIAStorage,
   GlacierInstantRetrievalStorage, GlacierStorage, DeepArchiveStorage).
2. Misplacement checks:
   a. Cold data still in STANDARD: archive/backup/compliance-named
      buckets with no lifecycle rules (get-bucket-lifecycle-configuration).
   b. Data in GLACIER (Flexible) that is never restored → cheaper in
      DEEP_ARCHIVE. Look for restore activity if request metrics exist.
   c. Data in GLACIER that gets restored urgently/often → belongs in
      GLACIER_IR.
   d. Lifecycle ladders that stop one rung early (e.g., transition to
      GLACIER but never to DEEP_ARCHIVE for multi-year retention).
3. Price ladder for the math (us-east-1, $/GB-mo): Standard 0.023,
   Standard-IA 0.0125, Glacier Instant 0.004, Glacier Flexible 0.0036,
   Deep Archive 0.00099. Minimum durations: 30/90/90/180 days.

Report: current GB and $/mo per class per bucket; each misplacement with
estimated monthly delta; a recommended target ladder per bucket as
lifecycle JSON — but do NOT apply anything.
Works with any assistant that can run shell commands.

Want the guided version?

The S3 Glacier (Flexible Retrieval) walkthrough covers this topic interactively — it asks about your setup, branches to what’s relevant, and quizzes you on the tricky parts. Free and anonymous.

Start the walkthrough

More Storage cheat sheets

AWS Backup Archive Tier

One lifecycle field on a backup plan moves long-retention recovery points to Glacier-backed cold storage at ~$1/TB-month — an 80–90% cut on backups whose realistic future is 'sits on a shelf until the retention clock runs out.'

EBS DeleteOnTermination

The wrong default on secondary EBS volumes leaves orphaned storage billing forever after instances terminate. One boolean in your launch templates stops the bleed; one audit command finds what already leaked.

EBS gp3 Volumes

Migrate EBS volumes from gp2 to gp3 for an immediate 20% price cut with the same or better performance — live, no downtime, one API call per volume.

EBS Snapshot Archive

Move compliance backups and 'just in case' EBS snapshots from Standard to the Archive tier for 75% lower storage cost — the trade is a 24–72 hour restore window that cold snapshots never actually need.