What 10 PB of Cold Data Really Costs in AWS, GCP, Azure vs Tape Over 20 Years

What a 10 PB, 20-year archive really costs in AWS/GCP/Azure vs a tape-backed object store Why long-term preservation is back on the board agenda Why long-term preservation is back on the board agenda Most cloud bills started life as rounding errors. A few terabytes here, a backup bucket there, a “just in case” archive for compliance. It looked cheap, flexible, and—critically—someone else’s problem. Fast-forward a few years: You’ve got petabytes sitting in “cold” tiers. Finance is staring at a 7-figure annual cloud line item for data that almost never changes. Legal and compliance are asking awkward questions about sovereignty, retention, and deletion guarantees. Sustainability is no longer a feel-good slide; it’s an emissions target with your company’s name on it. You’ve got petabytes sitting in “cold” tiers. petabytes Finance is staring at a 7-figure annual cloud line item for data that almost never changes. Legal and compliance are asking awkward questions about sovereignty, retention, and deletion guarantees. sovereignty, retention, and deletion guarantees Sustainability is no longer a feel-good slide; it’s an emissions target with your company’s name on it. This is where preservation storage (data you must keep for 5–20+ years) diverges from operational storage (data you actively use to run the business). Treat them the same, and you’ll overspend on one or under-protect the other. preservation storage operational storage Cloud cold storage is, in many ways, a rental model that works brilliantly for elastic workloads: elastic workloads Short-term experiments Bursty analytics Seasonal apps Disaster recovery copies you hope to never touch Short-term experiments Bursty analytics Seasonal apps Disaster recovery copies you hope to never touch But for regulatory archives, historical content, medical images, logs, and R&D datasets that must live for a decade or more, the economics and risk profile start to look very different. regulatory archives, historical content, medical images, logs, and R&D datasets That’s why repatriation—moving data out of the cloud and into an on-premise or hosted environment you control—is back in serious conversations. Not because “cloud was a mistake,” but because not all data belongs on a permanent subscription model. repatriation out not all data belongs on a permanent subscription model. This article walks through a concrete scenario: 10 petabytes of cold data, accessed about 2.5% per month, over 5 / 10 / 20 years, compared across major cloud cold tiers and a tape-based object store. 10 petabytes of cold data, accessed about 2.5% per month, over 5 / 10 / 20 years, compared across major cloud cold tiers and a tape-based object store. 10 petabytes of cold data, accessed about 2.5% per month, over 5 / 10 / 20 years tape-based object store We’ll keep the math visible, the tone honest, and the conclusion practical. The scenario: 10 PB under glass The scenario: 10 PB under glass Let’s pin down the workload so the numbers aren’t hand-wavy. Total archive size:10 PB = 10,000 TB ≈10,000,000 GB Access rate:2.5% of the data accessed per month→ 0.025 × 10,000,000 GB = 250,000 GB/month→ ~3,000,000 GB/year retrieved Total archive size:10 PB = 10,000 TB ≈10,000,000 GB Total archive size 10,000,000 GB Access rate:2.5% of the data accessed per month→ 0.025 × 10,000,000 GB = 250,000 GB/month→ ~3,000,000 GB/year retrieved Access rate 3,000,000 GB/year So you have: A large, relatively stable corpus (10 PB) Low but non-zero access (monthly reads for audits, investigations, restorations, etc.) A long horizon: 5, 10, 20 years A large, relatively stable corpus (10 PB) large, relatively stable corpus Low but non-zero access (monthly reads for audits, investigations, restorations, etc.) Low but non-zero access A long horizon: 5, 10, 20 years long horizon We’ll compare: Cloud cold storage classes AWS S3 Glacier Instant Retrieval S3 Glacier Flexible Retrieval S3 Glacier Deep Archive Google Cloud Platform Nearline Coldline Archive Azure Blob Storage Cool tier (Cool access tier) AWS S3 Glacier Instant Retrieval S3 Glacier Flexible Retrieval S3 Glacier Deep Archive AWS S3 Glacier Instant Retrieval S3 Glacier Flexible Retrieval S3 Glacier Deep Archive S3 Glacier Instant Retrieval S3 Glacier Flexible Retrieval S3 Glacier Deep Archive Google Cloud Platform Nearline Coldline Archive Google Cloud Platform Nearline Coldline Archive Nearline Coldline Archive Azure Blob Storage Cool tier (Cool access tier) Azure Blob Storage Cool tier (Cool access tier) Blob Storage Cool tier (Cool access tier) On-premise: a tape-backed object store Using a cost profile based on LTO.org-style numbers for a mid-size tape infrastructure: Hardware (Server): $35,000 Hardware (Disk storage / cache): $25,000 Hardware (Tape library + drives): $80,000 Media: $63,600 Software: $35,000 Hardware (Server): $35,000 $35,000 Hardware (Disk storage / cache): $25,000 $25,000 Hardware (Tape library + drives): $80,000 $80,000 Media: $63,600 $63,600 Software: $35,000 $35,000 → Initial CapEx = $238,600 Initial CapEx = $238,600 Ongoing: Operations staff: $5,000/month = $60,000/year Maintenance: $2,000/year Floor space: $4,500/month = $54,000/year, with 4% annual uplift Energy: not specified; we’ll treat it as zero in the math (real life: add maybe $10–15k/year). Operations staff: $5,000/month = $60,000/year $5,000/month $60,000/year Maintenance: $2,000/year $2,000/year Floor space: $4,500/month = $54,000/year, with 4% annual uplift $4,500/month $54,000/year 4% annual uplift Energy: not specified; we’ll treat it as zero in the math (real life: add maybe $10–15k/year). Refresh cycles: Server + disk: every 5 years → $60,000 at years 5, 10, 15, 20 Tape hardware + media: every 7 years → $143,600 at years 7, 14, 21 Server + disk: every 5 years → $60,000 at years 5, 10, 15, 20 every 5 years $60,000 Tape hardware + media: every 7 years → $143,600 at years 7, 14, 21 every 7 years $143,600 And we’ll add a one-time cloud egress cost to pull the full 10 PB home. cloud egress cost For egress we’ll assume a large-volume blended rate of $0.05/GB for cloud → on-prem transfers. Real discount deals will vary, but public pricing for big volumes lands in that neighborhood. $0.05/GB 10 PB = 10,000,000 GB → 10,000,000 GB × $0.05/GB ≈ $500,000 egress to repatriate. 10 PB = 10,000,000 GB → 10,000,000 GB × $0.05/GB ≈ $500,000 egress to repatriate. 10,000,000 GB × $0.05/GB ≈ $500,000 egress Cloud cold storage: the actual cost of “keeping it just in case” Cloud cold storage: the actual cost of “keeping it just in case” Let’s start with AWS, then we’ll look at GCP and Azure at a high level. Anatomy of cloud cold storage cost Anatomy of cloud cold storage cost For each cloud class, your annual cost has two main components: Storage rent: Storage rent: Storage rent: Annual storage cost=Data (GB)×Price ($/GB-month)×12 Annual storage cost=Data (GB)×Price ($/GB-month)×12 Retrieval + egress tolls: Retrieval + egress tolls: Retrieval + egress tolls Annual retrieval cost=Data retrieved annually (GB)×(retrieval fee+egress fee) Annual retrieval cost=Data retrieved annually (GB)×(retrieval fee+egress fee) In our case: Data stored: 10,000,000 GB Data retrieved annually: 3,000,000 GB Egress fee: $0.05/GB (assumed) Retrieval fee: varies by storage class Data stored: 10,000,000 GB 10,000,000 GB Data retrieved annually: 3,000,000 GB 3,000,000 GB Egress fee: $0.05/GB (assumed) $0.05/GB Retrieval fee: varies by storage class We’ll use typical US region prices (rounded) as of 2025 for illustration: AWS (US region; approximate base on AWS advertised pricing (your mileage may very)) S3 Glacier Instant Retrieval Storage: ~$0.004/GB-month Retrieval: ~$0.03/GB (class fee) S3 Glacier Flexible Retrieval Storage: ~$0.0036/GB-month Retrieval: ~$0.01/GB (standard retrieval) S3 Glacier Deep Archive Storage: $0.00099/GB-month (AWS public doc) Retrieval: ~$0.02/GB (typical deep archive retrieval) S3 Glacier Instant Retrieval Storage: ~$0.004/GB-month Retrieval: ~$0.03/GB (class fee) S3 Glacier Instant Retrieval Storage: ~$0.004/GB-month Retrieval: ~$0.03/GB (class fee) Storage: ~$0.004/GB-month $0.004/GB-month Retrieval: ~$0.03/GB (class fee) $0.03/GB S3 Glacier Flexible Retrieval Storage: ~$0.0036/GB-month Retrieval: ~$0.01/GB (standard retrieval) S3 Glacier Flexible Retrieval Storage: ~$0.0036/GB-month Retrieval: ~$0.01/GB (standard retrieval) Storage: ~$0.0036/GB-month $0.0036/GB-month Retrieval: ~$0.01/GB (standard retrieval) $0.01/GB S3 Glacier Deep Archive Storage: $0.00099/GB-month (AWS public doc) Retrieval: ~$0.02/GB (typical deep archive retrieval) S3 Glacier Deep Archive Storage: $0.00099/GB-month (AWS public doc) Retrieval: ~$0.02/GB (typical deep archive retrieval) Storage: $0.00099/GB-month (AWS public doc) $0.00099/GB-month Retrieval: ~$0.02/GB (typical deep archive retrieval) $0.02/GB GCP and Azure have similar shapes—slightly different list prices, but the same pattern: low storage, non-trivial retrieval and egress. non-trivial retrieval and egress Worked example: AWS S3 Glacier Deep Archive Worked example: AWS S3 Glacier Deep Archive Let’s fully show the math for the cheapest major cloud cold tier: S3 Glacier Deep Archive. the cheapest major cloud cold tier Inputs Inputs Data stored: 10,000,000 GB Storage price (S): $0.00099 per GB-month Annual retrieval volume: 3,000,000 GB Retrieval fee (R): $0.02 per GB Egress fee (E): $0.05 per GB Data stored: 10,000,000 GB Storage price (S): $0.00099 per GB-month Annual retrieval volume: 3,000,000 GB Retrieval fee (R): $0.02 per GB Egress fee (E): $0.05 per GB Step 1 – Annual storage cost Step 1 – Annual storage cost Annual storage=10,000,000×0.00099×12 Annual storage=10,000,000×0.00099×12 10,000,000 × 0.00099 = 9,900 9,900 × 12 = $118,800/year 10,000,000 × 0.00099 = 9,900 9,900 × 12 = $118,800/year $118,800/year Step 2 – Annual retrieval + egress cost Step 2 – Annual retrieval + egress cost Annual retrieval=3,000,000×(0.02+0.05) Annual retrieval=3,000,000×(0.02+0.05) Retrieval + egress = 0.02 + 0.05 = $0.07/GB 3,000,000 × 0.07 = $210,000/year Retrieval + egress = 0.02 + 0.05 = $0.07/GB $0.07/GB 3,000,000 × 0.07 = $210,000/year $210,000/year Step 3 – Total annual cost Step 3 – Total annual cost Annual total=118,800+210,000=$328,800 Annual total=118,800+210,000=$328,800 Step 4 – 5, 10, 20-year totals Step 4 – 5, 10, 20-year totals 5 years: 5 × 328,800 = $1,644,000 10 years: 10 × 328,800 = $3,288,000 20 years: 20 × 328,800 = $6,576,000 5 years: 5 × 328,800 = $1,644,000 5 years $1,644,000 10 years: 10 × 328,800 = $3,288,000 10 years $3,288,000 20 years: 20 × 328,800 = $6,576,000 20 years $6,576,000 So even in the cheapest AWS cold class, 10 PB with modest ongoing access is: cheapest AWS cold class ~$1.64M over 5 years ~$3.29M over 10 years ~$6.58M over 20 years ~$1.64M over 5 years $1.64M over 5 years ~$3.29M over 10 years $3.29M over 10 years ~$6.58M over 20 years $6.58M over 20 years …and that’s ignoring: Minimum storage duration penalties (90–180 days) Per-request fees Staging copies into S3 Standard while restores are “rehydrated” Minimum storage duration penalties (90–180 days) Per-request fees Staging copies into S3 Standard while restores are “rehydrated” You pay rent forever and tolls every time you read. rent forever tolls every time you read High-level look at other cloud cold tiers Using similar reasoning and current public price examples (storage + retrieval + egress), you end up in this ballpark for 10 PB and the same access pattern: Order-of-magnitude TCO for 10 PB @ 2.5%/month access Order-of-magnitude TCO for 10 PB @ 2.5%/month access Order-of-magnitude TCO for 10 PB @ 2.5%/month access All numbers rounded to nearest $0.1M, assuming similar $0.05/GB egress and representative retrieval fees. All numbers rounded to nearest $0.1M, assuming similar $0.05/GB egress and representative retrieval fees. AWS Glacier Instant: ~$3.6M (5 yr), $7.2M (10 yr), $14.4M (20 yr) Glacier Flexible: ~$3.1M, $6.1M, $12.2M Glacier Deep Archive: ~$1.6M, $3.3M, $6.6M GCP (Nearline / Coldline / Archive with typical prices like 0.010 / 0.007 / 0.004 $ per GB-month) Nearline: ~$6.9M, $13.8M, $27.6M Coldline: ~$5.3M, $10.5M, $21.0M Archive: ~$3.9M, $7.8M, $15.6M Azure Blob Cool (≈ $0.01–0.0115/GB-month plus retrieval + egress; Azure also has Cold/Archive tiers, but you asked specifically for Cool) Blob Cool: ~$7.8M, $15.6M, $31.2M AWS Glacier Instant: ~$3.6M (5 yr), $7.2M (10 yr), $14.4M (20 yr) Glacier Flexible: ~$3.1M, $6.1M, $12.2M Glacier Deep Archive: ~$1.6M, $3.3M, $6.6M AWS Glacier Instant: ~$3.6M (5 yr), $7.2M (10 yr), $14.4M (20 yr) Glacier Flexible: ~$3.1M, $6.1M, $12.2M Glacier Deep Archive: ~$1.6M, $3.3M, $6.6M Glacier Instant: ~$3.6M (5 yr), $7.2M (10 yr), $14.4M (20 yr) ~$3.6M (5 yr) $7.2M (10 yr) $14.4M (20 yr) Glacier Flexible: ~$3.1M, $6.1M, $12.2M ~$3.1M $6.1M $12.2M Glacier Deep Archive: ~$1.6M, $3.3M, $6.6M ~$1.6M $3.3M $6.6M GCP (Nearline / Coldline / Archive with typical prices like 0.010 / 0.007 / 0.004 $ per GB-month) Nearline: ~$6.9M, $13.8M, $27.6M Coldline: ~$5.3M, $10.5M, $21.0M Archive: ~$3.9M, $7.8M, $15.6M GCP Nearline: ~$6.9M, $13.8M, $27.6M Coldline: ~$5.3M, $10.5M, $21.0M Archive: ~$3.9M, $7.8M, $15.6M Nearline: ~$6.9M, $13.8M, $27.6M ~$6.9M $13.8M $27.6M Coldline: ~$5.3M, $10.5M, $21.0M ~$5.3M $10.5M $21.0M Archive: ~$3.9M, $7.8M, $15.6M ~$3.9M $7.8M $15.6M Azure Blob Cool (≈ $0.01–0.0115/GB-month plus retrieval + egress; Azure also has Cold/Archive tiers, but you asked specifically for Cool) Blob Cool: ~$7.8M, $15.6M, $31.2M Azure Blob Cool Blob Cool: ~$7.8M, $15.6M, $31.2M Blob Cool: ~$7.8M, $15.6M, $31.2M ~$7.8M $15.6M $31.2M The broad pattern: Anything “Instant” or “Cool”: easily $6–15M+ over 10–20 years. Deepest archive tiers (AWS Deep Archive, GCP Archive, Azure Archive) lower the storage part, but retrieval + egress still bite. Anything “Instant” or “Cool”: easily $6–15M+ over 10–20 years. Anything “Instant” or “Cool” $6–15M+ over 10–20 years Deepest archive tiers (AWS Deep Archive, GCP Archive, Azure Archive) lower the storage part, but retrieval + egress still bite. Tape-backed object store: your own cold cloud Tape-backed object store: your own cold cloud Now let’s look at the tape-backed object store using the numbers you provided. One-time CapEx One-time CapEx Server hardware: $35,000 Disk storage (cache): $25,000 Tape library + drives: $80,000 Tape media: $63,600 Software: $35,000 Server hardware: $35,000 Disk storage (cache): $25,000 Tape library + drives: $80,000 Tape media: $63,600 Software: $35,000 Initial CapEx=35,000+25,000+80,000+63,600+35,000=$238,600 Initial CapEx=35,000+25,000+80,000+63,600+35,000=$238,600 So $238,600 to stand up the environment. $238,600 Annual OpEx (baseline) Annual OpEx (baseline) Operations staff: $5,000/month → $60,000/year Maintenance: $2,000/year Floor space: $4,500/month → $54,000/year, with 4% annual increase Operations staff: $5,000/month → $60,000/year $5,000/month $60,000/year Maintenance: $2,000/year $2,000/year Floor space: $4,500/month → $54,000/year, with 4% annual increase $4,500/month $54,000/year 4% annual increase Ignoring energy for now, your year-1 baseline OpEx: year-1 baseline OpEx Year 1 OpEx=60,000+2,000+54,000=$116,000 Year 1 OpEx=60,000+2,000+54,000=$116,000 From year 2 onward, floor space increases: Floor space in year n=54,000×(1.04)^(n−1) Floor space in year n=54,000×(1.04)^(n−1) Everything else (ops + maintenance) we’ll keep flat for simplicity. Refresh cycles Refresh cycles Every 5 years (5, 10, 15, 20): $60,000 for server + disk upgrades Every 7 years (7, 14, 21…): $143,600 for tape hardware + new media (if needed) Every 5 years (5, 10, 15, 20): $60,000 for server + disk upgrades $60,000 Every 7 years (7, 14, 21…): $143,600 for tape hardware + new media (if needed) $143,600 These are big, but they happen only a few times over 20 years. Putting it together: TCO over 5, 10, 20 years Putting it together: TCO over 5, 10, 20 years Running the math with: Initial CapEx Annual OpEx (with escalating floor space) Scheduled refreshes Initial CapEx Annual OpEx (with escalating floor space) Scheduled refreshes You get approximate totals: 5-year TCO (tape env only) ≈ $0.90M 10-year TCO ≈ $1.77M 20-year TCO ≈ $3.61M 5-year TCO (tape env only) ≈ $0.90M 5-year TCO (tape env only) $0.90M 10-year TCO ≈ $1.77M 10-year TCO $1.77M 20-year TCO ≈ $3.61M 20-year TCO $3.61M Now add the one-time egress to bring the 10 PB home: one-time egress 10,000,000 GB × $0.05/GB = $500,000 10,000,000 GB × $0.05/GB = $500,000 10,000,000 GB × $0.05/GB = $500,000 So “tape + migration” TCO: “tape + migration” TCO: Horizon Tape env only + 10 PB egress Total Tape+Egress TCO 5yr $0.90M $0.50M $1.40M 10yr $1.77M $0.50M $2.27M 20yr $3.61M $0.50M $4.11M Horizon Tape env only + 10 PB egress Total Tape+Egress TCO 5yr $0.90M $0.50M $1.40M 10yr $1.77M $0.50M $2.27M 20yr $3.61M $0.50M $4.11M Horizon Tape env only + 10 PB egress Total Tape+Egress TCO Horizon Horizon Tape env only Tape env only + 10 PB egress + 10 PB egress Total Tape+Egress TCO Total Tape+Egress TCO 5yr $0.90M $0.50M $1.40M 5yr 5yr $0.90M $0.90M $0.50M $0.50M $1.40M $1.40M 10yr $1.77M $0.50M $2.27M 10yr 10yr $1.77M $1.77M $0.50M $0.50M $2.27M $2.27M 20yr $3.61M $0.50M $4.11M 20yr 20yr $3.61M $3.61M $0.50M $0.50M $4.11M $4.11M Remember: that’s with no energy cost added. Even if you add $10–15k/year for power and cooling, the totals barely move compared to multi-million cloud bills. no energy cost Head-to-head: cloud vs tape for 10 PB Head-to-head: cloud vs tape for 10 PB Versus AWS S3 Glacier Deep Archive Versus AWS S3 Glacier Deep Archive Compare “cheapest viable cloud cold” vs “your own tape cloud”: Horizon AWS S3 Glacier Deep Archive Tape + Egress Cloud – Tape (extra spend) 5yr $1.64M $1.4M +0.24M 10yr $3.29M $2.27M +1.02M 20yr $6.58M $4.11M +2.47M Horizon AWS S3 Glacier Deep Archive Tape + Egress Cloud – Tape (extra spend) 5yr $1.64M $1.4M +0.24M 10yr $3.29M $2.27M +1.02M 20yr $6.58M $4.11M +2.47M Horizon AWS S3 Glacier Deep Archive Tape + Egress Cloud – Tape (extra spend) Horizon Horizon AWS S3 Glacier Deep Archive AWS S3 Glacier Deep Archive Tape + Egress Tape + Egress Cloud – Tape (extra spend) Cloud – Tape (extra spend) 5yr $1.64M $1.4M +0.24M 5yr 5yr $1.64M $1.64M $1.4M $1.4M +0.24M +0.24M 10yr $3.29M $2.27M +1.02M 10yr 10yr $3.29M $3.29M $2.27M $2.27M +1.02M +1.02M 20yr $6.58M $4.11M +2.47M 20yr 20yr $6.58M $6.58M $4.11M $4.11M +2.47M +2.47M Interpretation: Over 5 years, tape wins, but by a modest margin (~$240k). Over 10 years, tape+egress is about $1M cheaper. Over 20 years, tape+egress is roughly $2.5M cheaper. Over 5 years, tape wins, but by a modest margin (~$240k). 5 years Over 10 years, tape+egress is about $1M cheaper. 10 years $1M cheaper Over 20 years, tape+egress is roughly $2.5M cheaper. 20 years $2.5M cheaper And that’s against the cheapest of the mainstream cloud cold tiers. cheapest Versus more expensive tiers (Instant, Nearline, Cool), tape wins by multiple millions. Pros and cons: this isn’t just about dollars Pros and cons: this isn’t just about dollars Cloud cold storage – PROs Cloud cold storage – PROs No CapEx, no capacity planning You “flip a switch,” you pay monthly. Good when future capacity needs are uncertain. Global durability and availability baked in Multi-AZ, multi-region options; you inherit their durability and uptime SLAs. Operational simplicity No hardware lifecycle management, no tape handling, no real estate decisions. Elastic recovery and DR use cases Easy to wire into cloud-native DR patterns, cross-region replication, etc. No CapEx, no capacity planning You “flip a switch,” you pay monthly. Good when future capacity needs are uncertain. No CapEx, no capacity planning You “flip a switch,” you pay monthly. Good when future capacity needs are uncertain. You “flip a switch,” you pay monthly. Good when future capacity needs are uncertain. Global durability and availability baked in Multi-AZ, multi-region options; you inherit their durability and uptime SLAs. Global durability and availability Multi-AZ, multi-region options; you inherit their durability and uptime SLAs. Multi-AZ, multi-region options; you inherit their durability and uptime SLAs. Operational simplicity No hardware lifecycle management, no tape handling, no real estate decisions. Operational simplicity No hardware lifecycle management, no tape handling, no real estate decisions. No hardware lifecycle management, no tape handling, no real estate decisions. Elastic recovery and DR use cases Easy to wire into cloud-native DR patterns, cross-region replication, etc. Elastic recovery and DR use cases Easy to wire into cloud-native DR patterns, cross-region replication, etc. Easy to wire into cloud-native DR patterns, cross-region replication, etc. Cloud cold storage – CONs Cloud cold storage – CONs Permanent rental model The longer you keep data, the more the cloud wins—not you. For 10–20-year archives, the TCO behaves like a perpetual tax. Retrieval + egress penalties You pay every time you actually use your preserved data. Harder to justify new analytics or reuse on old data when the meter is running. Pricing opacity and volatility New tiers, new knobs, “discounts” that add complexity. It’s easy to end up paying for the wrong tier or misconfigured lifecycle. Lock-in The more archive data you park, the harder it becomes politically and financially to leave. Permanent rental model The longer you keep data, the more the cloud wins—not you. For 10–20-year archives, the TCO behaves like a perpetual tax. Permanent rental model The longer you keep data, the more the cloud wins—not you. For 10–20-year archives, the TCO behaves like a perpetual tax. The longer you keep data, the more the cloud wins—not you. not you For 10–20-year archives, the TCO behaves like a perpetual tax. perpetual tax Retrieval + egress penalties You pay every time you actually use your preserved data. Harder to justify new analytics or reuse on old data when the meter is running. Retrieval + egress penalties You pay every time you actually use your preserved data. Harder to justify new analytics or reuse on old data when the meter is running. You pay every time you actually use your preserved data. Harder to justify new analytics or reuse on old data when the meter is running. Pricing opacity and volatility New tiers, new knobs, “discounts” that add complexity. It’s easy to end up paying for the wrong tier or misconfigured lifecycle. Pricing opacity and volatility New tiers, new knobs, “discounts” that add complexity. It’s easy to end up paying for the wrong tier or misconfigured lifecycle. New tiers, new knobs, “discounts” that add complexity. It’s easy to end up paying for the wrong tier or misconfigured lifecycle. Lock-in The more archive data you park, the harder it becomes politically and financially to leave. Lock-in The more archive data you park, the harder it becomes politically and financially to leave. The more archive data you park, the harder it becomes politically and financially to leave. Tape-backed object store – PROs Tape-backed object store – PROs TCO flattens over time After you absorb CapEx and migration, your run-rate is stable, and the 5/7-year refresh cycles are predictable. The 10–20-year graph looks more like a plateau than an ever-increasing slope. No per-GB retrieval fees You can read as much as you want; the cost is operational, not metered. Incentivizes more reuse of historical data instead of hoarding it in the dark. Control and sovereignty Physical control over media. You decide destruction procedures, air-gapping, and chain of custody. Preservation-friendly Well-designed tape workflows (with checksums, periodic scrubbing, and migration) are still considered among the most robust strategies for long-term digital preservation. TCO flattens over time After you absorb CapEx and migration, your run-rate is stable, and the 5/7-year refresh cycles are predictable. The 10–20-year graph looks more like a plateau than an ever-increasing slope. TCO flattens over time After you absorb CapEx and migration, your run-rate is stable, and the 5/7-year refresh cycles are predictable. The 10–20-year graph looks more like a plateau than an ever-increasing slope. After you absorb CapEx and migration, your run-rate is stable, and the 5/7-year refresh cycles are predictable. run-rate is stable The 10–20-year graph looks more like a plateau than an ever-increasing slope. plateau No per-GB retrieval fees You can read as much as you want; the cost is operational, not metered. Incentivizes more reuse of historical data instead of hoarding it in the dark. No per-GB retrieval fees You can read as much as you want; the cost is operational, not metered. Incentivizes more reuse of historical data instead of hoarding it in the dark. You can read as much as you want; the cost is operational, not metered. Incentivizes more reuse of historical data instead of hoarding it in the dark. Control and sovereignty Physical control over media. You decide destruction procedures, air-gapping, and chain of custody. Control and sovereignty Physical control over media. You decide destruction procedures, air-gapping, and chain of custody. Physical control over media. You decide destruction procedures, air-gapping, and chain of custody. Preservation-friendly Well-designed tape workflows (with checksums, periodic scrubbing, and migration) are still considered among the most robust strategies for long-term digital preservation. Preservation-friendly Well-designed tape workflows (with checksums, periodic scrubbing, and migration) are still considered among the most robust strategies for long-term digital preservation. Well-designed tape workflows (with checksums, periodic scrubbing, and migration) are still considered among the most robust strategies for long-term digital preservation. Tape-backed object store – CONs Tape-backed object store – CONs Up-front CapEx and project risk You need budget, design time, and internal champions to execute. Migration from cloud introduces scheduling, bandwidth, and cutover risk. Operational complexity Tape libraries, drives, media management, and monitoring are not “set and forget.” You need people (or managed services) with relevant skills. Physical dependency Libraries and robotics can fail. You must design for spare parts, vendor support windows, and disaster recovery. Scalability and agility trade-offs Scaling beyond initial design requires deliberate planning and additional CapEx. You don’t get the same “spin up another region” flexibility you have in cloud. Up-front CapEx and project risk You need budget, design time, and internal champions to execute. Migration from cloud introduces scheduling, bandwidth, and cutover risk. Up-front CapEx and project risk You need budget, design time, and internal champions to execute. Migration from cloud introduces scheduling, bandwidth, and cutover risk. You need budget, design time, and internal champions to execute. Migration from cloud introduces scheduling, bandwidth, and cutover risk. Operational complexity Tape libraries, drives, media management, and monitoring are not “set and forget.” You need people (or managed services) with relevant skills. Operational complexity Tape libraries, drives, media management, and monitoring are not “set and forget.” You need people (or managed services) with relevant skills. Tape libraries, drives, media management, and monitoring are not “set and forget.” You need people (or managed services) with relevant skills. Physical dependency Libraries and robotics can fail. You must design for spare parts, vendor support windows, and disaster recovery. Physical dependency Libraries and robotics can fail. You must design for spare parts, vendor support windows, and disaster recovery. Libraries and robotics can fail. You must design for spare parts, vendor support windows, and disaster recovery. Scalability and agility trade-offs Scaling beyond initial design requires deliberate planning and additional CapEx. You don’t get the same “spin up another region” flexibility you have in cloud. Scalability and agility trade-offs Scaling beyond initial design requires deliberate planning and additional CapEx. You don’t get the same “spin up another region” flexibility you have in cloud. Scaling beyond initial design requires deliberate planning and additional CapEx. You don’t get the same “spin up another region” flexibility you have in cloud. When does repatriation make sense? When does repatriation make sense? From a C-suite / IT management lens, you repatriate preservation data when: The dataset is large enough Order of magnitude: ≥ 5–10 PB and growing. Below that, the complexity may outweigh the savings unless there are strong non-financial drivers. The access pattern is low but steady You do touch the data (audits, research, investigations, training sets) but not daily. Enough to be penalized by retrieval/egress, but not enough to justify hot storage. The retention horizon is long 10+ years is where the cloud rent really starts to look ugly compared to “owning the house.” Regulatory or institutional mandates often live in this band (records, medical, scientific, cultural). Governance and sovereignty matter You need to prove, with confidence, that: Data hasn’t been altered. Retention and deletion policies are actually enforced. Data is stored in specific jurisdictions under your operational control. You can staff or source the operational competency Either in-house tape and archive expertise, or a managed service provider who can run the environment with clear SLAs. The dataset is large enough Order of magnitude: ≥ 5–10 PB and growing. Below that, the complexity may outweigh the savings unless there are strong non-financial drivers. The dataset is large enough Order of magnitude: ≥ 5–10 PB and growing. Below that, the complexity may outweigh the savings unless there are strong non-financial drivers. Order of magnitude: ≥ 5–10 PB and growing. ≥ 5–10 PB Below that, the complexity may outweigh the savings unless there are strong non-financial drivers. The access pattern is low but steady You do touch the data (audits, research, investigations, training sets) but not daily. Enough to be penalized by retrieval/egress, but not enough to justify hot storage. The access pattern is low but steady You do touch the data (audits, research, investigations, training sets) but not daily. Enough to be penalized by retrieval/egress, but not enough to justify hot storage. You do touch the data (audits, research, investigations, training sets) but not daily. do Enough to be penalized by retrieval/egress, but not enough to justify hot storage. The retention horizon is long 10+ years is where the cloud rent really starts to look ugly compared to “owning the house.” Regulatory or institutional mandates often live in this band (records, medical, scientific, cultural). The retention horizon is long 10+ years is where the cloud rent really starts to look ugly compared to “owning the house.” Regulatory or institutional mandates often live in this band (records, medical, scientific, cultural). 10+ years is where the cloud rent really starts to look ugly compared to “owning the house.” 10+ years Regulatory or institutional mandates often live in this band (records, medical, scientific, cultural). Governance and sovereignty matter You need to prove, with confidence, that: Data hasn’t been altered. Retention and deletion policies are actually enforced. Data is stored in specific jurisdictions under your operational control. Governance and sovereignty matter You need to prove, with confidence, that: Data hasn’t been altered. Retention and deletion policies are actually enforced. Data is stored in specific jurisdictions under your operational control. You need to prove, with confidence, that: Data hasn’t been altered. Retention and deletion policies are actually enforced. Data is stored in specific jurisdictions under your operational control. Data hasn’t been altered. Retention and deletion policies are actually enforced. Data is stored in specific jurisdictions under your operational control. Data hasn’t been altered. Retention and deletion policies are actually enforced. Data is stored in specific jurisdictions under your operational control. You can staff or source the operational competency Either in-house tape and archive expertise, or a managed service provider who can run the environment with clear SLAs. You can staff or source the operational competency Either in-house tape and archive expertise, or a managed service provider who can run the environment with clear SLAs. Either in-house tape and archive expertise, or a managed service provider who can run the environment with clear SLAs. If you check most of those boxes, cloud cold storage for preservation becomes less “strategic agility” and more “expensive comfort blanket.” A practical decision framework for the board and CIO A practical decision framework for the board and CIO If you’re trying to decide what to do with a 10 PB archive sitting in Glacier/Coldline/Blob Cool, here’s a simple roadmap you can put in front of the C-suite. Step 1 – Inventory and segment the archive Step 1 – Inventory and segment the archive How much of the 10 PB is: Regulatory / compliance data? High-value institutional memory (R&D, design, media, logs)? Low-value “just in case” junk? Segment by: Retention requirement (years) Access pattern (frequency, type of queries) Business owner and risk tolerance How much of the 10 PB is: Regulatory / compliance data? High-value institutional memory (R&D, design, media, logs)? Low-value “just in case” junk? Regulatory / compliance data? High-value institutional memory (R&D, design, media, logs)? Low-value “just in case” junk? Regulatory / compliance data? High-value institutional memory (R&D, design, media, logs)? Low-value “just in case” junk? Segment by: Retention requirement (years) Access pattern (frequency, type of queries) Business owner and risk tolerance Retention requirement (years) Access pattern (frequency, type of queries) Business owner and risk tolerance Retention requirement (years) Retention requirement Access pattern (frequency, type of queries) Access pattern Business owner and risk tolerance Business owner Step 2 – Build a real baseline of current cloud cost Step 2 – Build a real baseline of current cloud cost Pull 12–24 months of actual bills. Separate: Storage charges (by tier) Retrieval charges Egress charges Per-request charges Project those costs over 5 / 10 / 20 years using your actual growth assumptions, not vendor marketing examples. Pull 12–24 months of actual bills. 12–24 months Separate: Storage charges (by tier) Retrieval charges Egress charges Per-request charges Storage charges (by tier) Retrieval charges Egress charges Per-request charges Storage charges (by tier) Retrieval charges Egress charges Per-request charges Project those costs over 5 / 10 / 20 years using your actual growth assumptions, not vendor marketing examples. 5 / 10 / 20 years This is where the 10 PB, 2.5% monthly access math above gives you a sanity check: if your model says your TCO will be $800k over 20 years, the model—not the math—is wrong. sanity check Step 3 – Design an on-prem (or hosted) preservation tier Step 3 – Design an on-prem (or hosted) preservation tier Size the tape environment (drives, slots, media) and disk cache for: Current 10 PB + expected growth. Monthly access pattern (enough drives/cache to keep SLA reasonable). Get real quotes for: Hardware (library, drives, servers, storage) Software (object storage, HSM, preservation workflows) Support contracts Floor space and power Size the tape environment (drives, slots, media) and disk cache for: Current 10 PB + expected growth. Monthly access pattern (enough drives/cache to keep SLA reasonable). Current 10 PB + expected growth. Monthly access pattern (enough drives/cache to keep SLA reasonable). Current 10 PB + expected growth. Monthly access pattern (enough drives/cache to keep SLA reasonable). Get real quotes for: Hardware (library, drives, servers, storage) Software (object storage, HSM, preservation workflows) Support contracts Floor space and power real quotes Hardware (library, drives, servers, storage) Software (object storage, HSM, preservation workflows) Support contracts Floor space and power Hardware (library, drives, servers, storage) Software (object storage, HSM, preservation workflows) Support contracts Floor space and power Compare the multi-year CapEx/OpEx profile against your cloud baseline. Step 4 – Plan the migration and cutover Step 4 – Plan the migration and cutover Decide how quickly you want to move 10 PB out: Network only? Cloud provider’s bulk export devices? Hybrid (start with hottest subsets)? Take into account: Network capacity and impact on other workloads. Order of operations (which buckets/collections first). Parallel validation and fixity checking (hash verification). Temporary storage needed for staging. Decide how quickly you want to move 10 PB out: Network only? Cloud provider’s bulk export devices? Hybrid (start with hottest subsets)? how quickly Network only? Cloud provider’s bulk export devices? Hybrid (start with hottest subsets)? Network only? Cloud provider’s bulk export devices? Hybrid (start with hottest subsets)? Take into account: Network capacity and impact on other workloads. Order of operations (which buckets/collections first). Parallel validation and fixity checking (hash verification). Temporary storage needed for staging. Network capacity and impact on other workloads. Order of operations (which buckets/collections first). Parallel validation and fixity checking (hash verification). Temporary storage needed for staging. Network capacity and impact on other workloads. Order of operations (which buckets/collections first). Parallel validation and fixity checking (hash verification). Temporary storage needed for staging. This is where the one-time $500k egress is a financial decision: pay now, buy back long-term control. $500k egress Step 5 – Embed governance and preservation practices Step 5 – Embed governance and preservation practices Regardless of where the data lives, for long-term preservation you need: Clear retention schedules and legal holds. Fixity checks (hash verification) over time. Format migration plans (file formats, codecs, schemas will age). Auditability (who accessed, what was restored, what was changed). Clear retention schedules and legal holds. retention schedules Fixity checks (hash verification) over time. Fixity checks Format migration plans (file formats, codecs, schemas will age). Format migration Auditability (who accessed, what was restored, what was changed). Auditability A tape-backed object store can become your institutional memory tier, but only if governance and process are designed along with the hardware. institutional memory tier Step 6 – Revisit every 5 years Step 6 – Revisit every 5 years On each tech refresh cycle (5 or 7 years), reassess: Storage media roadmap (LTO generations, new cold media technologies). Cloud pricing changes (cloud isn’t standing still). Regulatory changes (retention, privacy, jurisdiction). Sustainability targets. On each tech refresh cycle (5 or 7 years), reassess: Storage media roadmap (LTO generations, new cold media technologies). Cloud pricing changes (cloud isn’t standing still). Regulatory changes (retention, privacy, jurisdiction). Sustainability targets. Storage media roadmap (LTO generations, new cold media technologies). Cloud pricing changes (cloud isn’t standing still). Regulatory changes (retention, privacy, jurisdiction). Sustainability targets. Storage media roadmap (LTO generations, new cold media technologies). Cloud pricing changes (cloud isn’t standing still). Regulatory changes (retention, privacy, jurisdiction). Sustainability targets. The decision you make today is not irrevocable; what matters is not sleepwalking into paying millions in rent for what could have been an owned, well-run archival estate. millions in rent owned, well-run archival estate The bottom line for the C-suite The bottom line for the C-suite For operational workloads, cloud is still a powerful tool. For 10 PB of long-term preservation data with modest ongoing access, cloud cold tiers look a lot like a lifetime subscription to your own history, with a toll booth every time you try to read it. 10 PB of long-term preservation data with modest ongoing access lifetime subscription A tape-backed object store, properly designed and operated, turns that into: A predictable, flattening TCO curve over 20 years. No per-GB read penalties, encouraging reuse and analysis. Improved sovereignty, governance, and preservation control. A predictable, flattening TCO curve over 20 years. TCO curve No per-GB read penalties, encouraging reuse and analysis. No per-GB read penalties Improved sovereignty, governance, and preservation control. sovereignty, governance, and preservation control The decision is not “cloud bad, tape good.” It’s: Put fast, changeable, time-sensitive workloads where elasticity and rich services win (cloud). Put large, slow-changing, must-keep-for-decades collections where long-term economics and control win (tape-backed preservation tier). Put fast, changeable, time-sensitive workloads where elasticity and rich services win (cloud). Put large, slow-changing, must-keep-for-decades collections where long-term economics and control win (tape-backed preservation tier). Put fast, changeable, time-sensitive workloads where elasticity and rich services win (cloud). Put large, slow-changing, must-keep-for-decades collections where long-term economics and control win (tape-backed preservation tier). If you’re holding 10 PB of “cold” data in the cloud today and expecting to keep it there for 10–20 years, you’re not just storing history—you’re renting it. The question for the C-suite is simple: Do you want to keep paying rent forever, or do you want to own the building where your institutional memory lives? Do you want to keep paying rent forever, or do you want to own the building where your institutional memory lives? Do you want to keep paying rent forever, or do you want to own the building where your institutional memory lives?