An overwhelming majority (87%) of SQL restore events have their roots in human error, new research has discovered.

The most common human error is accidental deletion, accounting for one in three (35%) restore requests. This is followed by development errors (28%), erroneous overwrites (24%) and system crashes (13%) adding hours of recovery time for SQL database administrators (DBAs).

Further, respondents to the research by Kroll Ontrack said they field nearly twice as many requests to restore data to development and test environments compared to production environments.

This is a logical outcome considering that developers are continually trying out new things in development environments, making them highly iterative and therefore less stable.

>See also: The definitive guide to the modern database

In fact, on a monthly basis, 50% of respondents handle between one and five development and test restore requests, nearly 20% handle between six and ten requests, and 15% tackle more than ten requests each month.

With previous Kroll Ontrack survey results reporting database table restorations taking upwards of an hour to complete, leveraging efficient technologies to reduce restore time is key to saving time and money.

“It’s no secret that data housed within SQL databases is critically important to the business success of an organisation,” said Paul Le Messurier, programme and operations manager at Kroll Ontrack. “As such, any disruption to the database is inherently costly, putting pressure on DBAs to execute an efficient recovery solution.”