Backblaze Q2 2026 drive stats: failure rate climbs to 1.73% as aging drives drag on the fleet

Backblaze's Q2 2026 drive stats report puts the quarterly annualized failure rate at 1.73% across a fleet of 354,415 hard drives, the highest figure the company has recorded in some time. The cloud backup provider attributes the rise to an increasingly elderly fleet rather than a wave of bad hardware, while deployments continue to shift toward 20TB-plus capacities.

The quarter in numbers

The reporting period runs from April 1 to June 30, 2026. Backblaze monitored 359,101 drives in total, then set aside 3,881 boot drives and 705 drives that did not meet its inclusion criteria. A model needs more than 100 drives and more than 10,000 drive days to appear in the quarterly table. The remaining 354,415 drives logged roughly 31.6 million drive days and 1,498 failures.

According to Backblaze, the spread of failure rates is as notable as the overall figure: 10 of the 31 models in the quarterly table posted annualized failure rates above 3.0%. Three of those qualified as statistical outliers under the report's quartile analysis, which put the outlier threshold at 6.95% for this quarter:

• HGST HUH721212ALN604 (12TB) at 7.63%
• Seagate ST10000NM0086 (10TB) at 9.33%
• Seagate ST14000NM0138 (14TB) at 8.26%

Old drives and small samples

Backblaze ties each outlier to age or sample size. The HGST model is nearly seven years old on average, and the 10TB Seagate is around eight and a half. The two Seagate models also run in small populations of 965 and 1,235 units, so a modest number of failures, 22 and 25 respectively, produces a steep annualized rate.

Seagate also supplied the quarter's best results. Three of its models recorded zero failures: the 8TB ST8000NM000A, the 12TB ST12000NM000J and the 14TB ST14000NM000J. Its 16TB ST16000NM000J failed exactly once.

The fleet keeps consolidating at high capacity

Two long-serving models left the roster this quarter: the 4TB HGST HMS5C4040BLE640 at nearly nine years of service, and the 8TB HGST HUH728080ALE600 at nearly eight. For the second consecutive quarter no new drive models joined the fleet, but Backblaze notes that fresh deployment is concentrated almost entirely in drives of 20TB and above.

The early returns on those large drives are largely encouraging. WDC's 22TB WUH722222ALE6L4, now 45,665 units strong at an average age of 19.1 months, ran at 0.55%. Its 26TB sibling, the WUH722626ALE6L4, posted 0.95% across 7,212 drives averaging just 3.4 months in service. Toshiba's 24TB MG11ACA24TE recorded 0.58% for the quarter and 0.5% over its lifetime across 9,620 drives. The exception among the big-capacity cohort is Seagate's 24TB ST24000NM002H, at 3.3% for the quarter and 3.24% lifetime across 9,631 drives.

Lifetime figures elsewhere remain low. WDC's 14TB WUH721414ALE6L4 sits at 0.54%, and Seagate's 16TB ST16000NM001G at 0.68% across 35,100 drives.

What comes next

Beyond the tables, the report signals that Backblaze is preparing to compare its current CMR-based drives with newer high-capacity drive technology as those models take on a growing share of storage duties. The company is also hosting a review of the quarter's trends on October 8. As usual, the complete line-item dataset, including drives excluded from the article's tables, is available for download from Backblaze.

Why it matters

Backblaze has now published drive reliability data for more than 13 years, and it remains one of the very few public, large-scale datasets on hard drive failure. Storage buyers, data center operators and researchers use it to inform procurement and to sanity-check vendor claims. This quarter is a useful lesson in how to read it: fleet-wide averages drift upward as hardware ages, small drive populations can make individual models look far worse than they are, and the migration toward ever-denser drives, and eventually new recording technologies, is where the reliability story will be decided next.