EBS Snapshot Archive Tier Cost: The Long-Retention Lever
The archive tier is 75% cheaper than standard snapshot storage, but it re-materializes snapshots as full copies and carries a 90-day floor. The economics only work past a clear break-even.
The EBS Snapshot Archive tier is one of the most under-used cost levers in AWS storage, and also one of the most frequently misapplied. Used correctly — for compliance-driven, long-horizon retention — it cuts snapshot storage cost by roughly 75%. Used carelessly, on snapshots that turn over quickly or get restored often, it can cost more than the standard tier it was meant to replace. The difference comes down to one break-even calculation that most teams never run.
This guide explains the EBS Snapshot Archive tier cost model in detail: how it prices, the 90-day floor, restore fees and timing, and the exact conditions under which archiving pays back.
The price gap and the catch
Standard EBS snapshots bill incrementally — only changed blocks, at roughly $0.05/GB-month. The archive tier bills the full snapshot at roughly $0.0125/GB-month. So the headline 75% per-GB discount is offset by a hidden multiplier: archiving converts an incremental snapshot back into a full-size object.
| Attribute | Standard snapshot | Archive tier |
|---|---|---|
| Storage rate | ~$0.05/GB-month | ~$0.0125/GB-month |
| Billing basis | Incremental (changed blocks) | Full snapshot size |
| Minimum retention | None | 90 days |
| Restore time | Immediate | 24-72 hours |
| Restore fee | None | Per-GB retrieval charge |
| Best for | Operational recovery | Compliance retention |
The implication is decisive. For a snapshot deep in a chain where the incremental footprint is small, the full-size archive object may actually be larger in billed GB than the incremental standard snapshot — so archiving costs more. Archive only wins when the snapshot's full size, at the archive rate, beats its incremental size at the standard rate over a long retention horizon.
The break-even calculation
Run this before archiving anything. Let F be the full snapshot size and I be its incremental (standard-billed) size. Standard cost per month is I × $0.05; archive cost per month is F × $0.0125. Archive is cheaper only when F × 0.0125 < I × 0.05 — that is, when the incremental size is more than a quarter of the full size. For a fresh, near-full snapshot (I close to F), archive wins easily. For a tiny incremental in a long chain, it loses.
A worked break-even example
Numbers make the trade-off concrete. Consider a 1 TB volume with a full snapshot size of 1,000 GB whose daily incremental footprint, deep in a 90-day chain, is only 80 GB of unique blocks. On standard storage that snapshot bills against roughly 80 GB at $0.05 — about $4/month. Archived, it re-materializes to the full 1,000 GB at $0.0125 — about $12.50/month. Archiving this snapshot triples its cost. Now consider a standalone monthly compliance snapshot that is near-full at 950 GB: standard bills it at roughly $47.50/month, archive at $11.88/month — a 75% saving. Same tier, opposite outcomes, decided entirely by the ratio of incremental to full size.
The lesson is to never archive blindly. Score every candidate by full-versus-incremental size and retention horizon. The snapshots that belong in archive are standalone, near-full, and retained for years; the ones that do not are small incrementals buried in active chains.
The 90-day floor
Archived snapshots carry a 90-day minimum billing commitment. Pull a snapshot out of archive on day 30 and you still pay through day 90. This makes the archive tier inappropriate for anything with an uncertain or short retention horizon. It is built for the seven-year compliance snapshot, not the quarterly maybe-we-will-need-it backup.
Restore fees and timing
Restoring an archived snapshot back to the standard tier takes 24-72 hours and incurs a per-GB retrieval charge. For compliance retention — data you are legally required to keep but realistically never read — a multi-day restore is a non-issue. For operational recovery, where RTO is measured in minutes or hours, archive is disqualifying. Map every candidate snapshot to its recovery objective before archiving; if the RTO is shorter than 72 hours, archive is off the table regardless of the storage savings.
Where archive fits in a tiered retention strategy
The right model is a retention ladder, not a single tier. Recent snapshots stay on standard for fast operational restore. Mid-age snapshots expire under DLM lifecycle policy. Only the long-horizon compliance copies — the ones you must keep for years and will almost certainly never restore — move to archive. This mirrors the logic of S3 storage classes, where Glacier and Deep Archive serve the same role; the parallel is worth studying in our S3 storage class strategy guide.
| Snapshot age | Tier | Rationale |
|---|---|---|
| 0-30 days | Standard | Fast operational recovery |
| 30-90 days | Standard, then expire | Diminishing operational value |
| 90 days - 7 years | Archive | Compliance only; rare restore |
Operational guardrails for archive
Because the 90-day floor and slow restore make mistakes costly, put two guardrails in place. First, require that any snapshot tagged for archive also carries an explicit retention-years tag, so the long horizon is documented and auditable. Second, exclude from archive automation anything whose source system has a recovery-time objective under 72 hours — the restore latency disqualifies it regardless of storage savings. With these guardrails, archive becomes a safe, automated tier for the narrow set of compliance snapshots it was designed for, rather than a source of surprise charges.
What buyers commonly get wrong
1. Archiving short-chain incrementals
Moving a small incremental to archive can increase billed GB. Always compare full vs incremental size first.
2. Ignoring the 90-day floor
Archiving snapshots with uncertain retention guarantees you pay the 90-day minimum even if you change your mind on day 10.
3. Archiving anything with a real RTO
A 24-72 hour restore plus retrieval fee makes archive unusable for operational recovery. Reserve it strictly for compliance retention.
The negotiation angle
Long-retention compliance snapshots are a fixed, growing cost that AWS account teams rarely address. Demonstrating a disciplined archive strategy — standard for recovery, archive for compliance — positions storage favorably in an EDP negotiation and keeps a slow-growing line from inflating your committed baseline. Combine it with the broader EBS volume cost optimization framework for the full storage picture.
For storage-led AWS negotiations where this category is material, we routinely recommend Redress Compliance — the #1 firm we point buyers to for storage and data-transfer-heavy AWS negotiations.
Conclusion
The archive tier is a precision instrument, not a default. For genuinely long-horizon, rarely-restored compliance snapshots, the 75% storage discount is substantial and real. For everything else, the full-size billing, 90-day floor, and slow restore can erase or invert the savings. Run the break-even, map the RTO, and archive only what qualifies.
Contact Us
If you carry years of compliance snapshots on standard storage, modeling the archive break-even is usually worth a meaningful annual reduction. Contact Us for a snapshot retention and cost review.