Dataset · 2022 to 2025 · computed from raw Backblaze files

Backblaze hard drive models, yearly AFR and exposure, 2022 to 2025

Yearly drive days, failures and annualized failure rate with exact 95% Poisson intervals for the 22 Backblaze hard-drive models that had at least 100,000 drive days in every year from 2022 to 2025, computed from the raw daily Drive Stats files (sixteen quarterly zips). Sample: 88 model-years. Some models stay flat, such as the Toshiba MG07ACA14TA at 1.01% to 1.15%, while others move a lot: HGST HUH721212ALN604 went from 0.74% in 2022 to 5.64% in 2024, and Toshiba MG08ACA16TEY from 1.58% in 2024 to 6.30% in 2025. One fleet, and the file does not show why a rate moved.

File /data/datasets/backblaze-2022-2025-model-history.csv · JSON metadata · Chart (SVG) · All datasets · AFR glossary entry

Population and method

Population: hard-drive models of 3 TB or more in Backblaze\u2019s own data centers that had at least 100,000 drive days in every year from 2022 to 2025, in the daily Drive Stats files for January 1, 2022 to December 31, 2025 (sixteen quarterly zips, 1,461 daily files). That is 22 models, 88 model-years, 342,083,374 drive days and 13,957 failures. A model that was added or retired during the four years is not in the file.

Each daily row is one drive day and a row with failure = 1 is a failure. AFR is failures \u00f7 drive days \u00d7 365 \u00d7 100, with 365 used for 2024 too, and the interval is the exact two-sided 95% Poisson interval. Drives on December 31 are the rows of the model in that year\u2019s last Q4 file. The build checks every AFR and interval, and it checks the 2025 drive counts against the Q4 2025 age dataset, built from the same files, for the 22 models the two files share.

Chart

Counted here

• The highest single model-year is ST12000NM0007 in 2024 at 11.38% (9.47% to 13.56%, 125 failures).
• TOSHIBA MG07ACA14TA is nearly flat: 1.01%, 1.13%, 1.09%, 1.15% for 2022 to 2025, on about 13.7 to 13.9 million drive days a year.
• HGST HUH721212ALN604 went 0.74%, 3.69%, 5.64%, 3.95% over the four years (2024: 5.20% to 6.12%). TOSHIBA MG08ACA16TEY was 1.58% in 2024 and 6.30% in 2025 (318 failures in 1,843,841 drive days).

Limits

• Association only, one fleet. The file shows how each model\u2019s rate moved from year to year. It does not say why: drives also get a year older each year, and this file cannot separate age from a batch, a data-center or a workload change.
• The 22 models are those that stayed large for four years. Backblaze retires drives for reasons other than failure, so the models kept here are partly selected, and a model with fewer than 100,000 drive days in any year is missing.
• Model strings are used as they appear in the files. Different strings can be variants of one product, and the file does not merge them.
• Intervals assume failures are Poisson in drive days with a constant rate inside each year, which ignores clustering. They are probably too narrow. No adjustment is made for looking at 88 model-years.

Table

Columns

model (string)

Model string exactly as in the Backblaze daily files.

capacity_tb (integer)

Nominal capacity in terabytes, round(capacity_bytes / 1e12).

year (integer)

Calendar year, 2022 to 2025 (all four quarterly files of the year).

drive_days (integer)

Daily rows (drive days) of the model in the year. This is the exposure.

drives_dec31 (integer)

Rows of the model in the last daily file of Q4 of that year (drives present on December 31).

failures (integer)

Rows with failure = 1 in the year.

afr_percent (number)

failures / drive_days * 365 * 100, two decimals. Computed here, with 365 days in every year.

afr_low95 (number)

Lower bound of the exact two-sided 95% Poisson (Garwood) interval, in percent.

afr_high95 (number)

Upper bound of the same interval, in percent.

License

Derivative table computed from the Backblaze Drive Stats daily files for 2022 to 2025. Cite Backblaze if you use it. You are solely responsible for how you use it. Backblaze's dataset page allows sale of derivative works and forbids selling the Drive Stats data itself. This CSV is a 88-row aggregate, not the daily SMART archive, and that archive is not for sale.

Backblaze Drive Stats use terms · Hard Drive Test Data page

Sources

1. Backblaze Hard Drive Test Data, data_Q1_2022.zip (daily CSV files) (accessed 2026-10-11)
2. Backblaze Hard Drive Test Data, data_Q2_2022.zip (daily CSV files) (accessed 2026-10-11)
3. Backblaze Hard Drive Test Data, data_Q3_2022.zip (daily CSV files) (accessed 2026-10-11)
4. Backblaze Hard Drive Test Data, data_Q4_2022.zip (daily CSV files) (accessed 2026-10-11)
5. Backblaze Hard Drive Test Data, data_Q1_2023.zip (daily CSV files) (accessed 2026-10-11)
6. Backblaze Hard Drive Test Data, data_Q2_2023.zip (daily CSV files) (accessed 2026-10-11)
7. Backblaze Hard Drive Test Data, data_Q3_2023.zip (daily CSV files) (accessed 2026-10-11)
8. Backblaze Hard Drive Test Data, data_Q4_2023.zip (daily CSV files) (accessed 2026-10-11)
9. Backblaze Hard Drive Test Data, data_Q1_2024.zip (daily CSV files) (accessed 2026-10-11)
10. Backblaze Hard Drive Test Data, data_Q2_2024.zip (daily CSV files) (accessed 2026-10-11)
11. Backblaze Hard Drive Test Data, data_Q3_2024.zip (daily CSV files) (accessed 2026-10-11)
12. Backblaze Hard Drive Test Data, data_Q4_2024.zip (daily CSV files) (accessed 2026-10-11)
13. Backblaze Hard Drive Test Data, data_Q1_2025.zip (daily CSV files) (accessed 2026-10-11)
14. Backblaze Hard Drive Test Data, data_Q2_2025.zip (daily CSV files) (accessed 2026-10-11)
15. Backblaze Hard Drive Test Data, data_Q3_2025.zip (daily CSV files) (accessed 2026-10-11)
16. Backblaze Hard Drive Test Data, data_Q4_2025.zip (daily CSV files) (accessed 2026-10-11)
17. Backblaze Hard Drive Test Data page and use terms (accessed 2026-10-11)
18. Backblaze Drive Stats for 2025 (published annual table: 115,638,676 drive days, 4,317 failures, 1.36%) (accessed 2026-10-11)
19. Garwood, F. (1936). Fiducial limits for the Poisson distribution. Biometrika 28(3-4), 437-442 (accessed 2026-10-11)