Since 2013, Backblaze has collected, curated, and published the annualized failure rates (AFR) and related statistics from the hard disk drives (HDDs) and solid state drives (SSDs) in our data centers. This collection is the Backblaze Drive Stats dataset. Each quarter we publish updates to the dataset which is open source and can be downloaded using the links in the “Downloading the Drive Stats Dataset” section below.
Drive Stats Q2 2025 Snapshot
Drive count
317,230
Drive failures
1,061
Drive days
28,402,627
Drive population by manufacturer
HGST
Seagate
Toshiba
WDC
Drive reliability: annualized failure rates (AFR)
Drive Stats quarterly reports and related articles
We publish our analyses, observations, and insights based on the Drive Stats dataset on a regular basis on the Backblaze Blog which includes the quarterly Hard Drive Stats reports and SSD reports, and related topics such as the cost of storage, bathtub curve, enterprise-level drive management in our data centers, and more.
Overview of the Drive Stats dataset
How we collect the drive data
Every day at Backblaze data center, we take a snapshot of each operational drive. This snapshot includes basic drive information along with the S.M.A.R.T. statistics reported by that drive. The daily snapshot of one drive is one record or row of data. All of the drive snapshots for a given day are collected into a CSV file consisting of a row for each active drive. Each day this file is named in the format YYYY-MM-DD.csv; for example, 2024-03-25.csv.
How the Drive Stats data is organized
The Drive Stats schema is comprised of fields Backblaze includes for each drive record and the raw and normalized S.M.A.R.T. attributes reported by each drive.
- Download a CSV file of the current Drive Stats schema: https://f001.backblazeb2.com/file/Backblaze-Hard-Drive-Data/Drive_Stats_Schema_Current.csv
- Download a .csv file of the Drive Stats schema changes from Q1 2018 onward: https://f001.backblazeb2.com/file/Backblaze-Hard-Drive-Data/Drive_Stats_Schema_2018_Onward.csv
Please note, schema changes from quarter to quarter do occur, so you should always check for such changes each quarter and align the data to reflect any changes.
How you can use the Drive Stats data
The Drive Stats dataset is open source and available for you to download below, all we ask is is the following:
- Cite Backblaze as the source if you use the data,
- Accept that you are solely responsible for how you use the data,
- You may sell derivative works based on the data, but
- You can not sell the data itself to anyone; it is free.
Querying the Drive Stats dataset
As well as the compressed CSV files listed below, we maintain a copy of the Drive Stats dataset in Apache Iceberg table format. You should be able to use the following read-only credentials to query the dataset from any tools that support Apache Iceberg:
Application Key ID (AWS Key ID): 0045f0571db506a0000000017
Application Key (AWS Secret Key): K004Fs/bgmTk5dgo6GAVm2Waj3Ka+TE
Endpoint URL: https://s3.us-west-004.backblazeb2.com
Region: us-west-004
Bucket: drivestats-iceberg
Path Prefix: drivestats
See the blog post Iceberg on Backblaze B2 for comprehensive instructions on how to directly query the Drive Stats dataset from Trino, Snowflake, and DuckDB.
Downloading the Backblaze Drive Stats dataset
Beginning in 2016 we uploaded the Drive Stats dataset for a given quarter. Prior to 2016 the datasets uploaded were annual (2013, 2014, and 2015). Each item listed below is a ZIP file of containing the .csv files for the named quarter or year.
