Disk fault prediction method and system, terminal and storage medium

Abstract

The invention provides a disk fault prediction method and system, a terminal and a storage medium. The disk fault prediction method comprises the steps of screening SMART feature items related to diskfault prediction by utilizing a recursive feature elimination algorithm; generating a disk health parameter according to the fluctuation condition of the related feature item of the disk in the preset time period; training a neural network model by using the historical disk SMART feature data; and inputting the disk health parameters into a neural network model to obtain a disk fault prediction result. The SMART features are screened by using the RFE algorithm, the number of the features is reduced, the calculated amount is reduced, overfitting is inhibited, the new feature value is generatedaccording to the change times of the SMART feature value within seven days, and the proportion, the balance accuracy and the recall rate of two types of errors in the loss function are trained and adjusted more easily.

Description

technical field [0001] The technical field of the big data service platform of the present invention specifically relates to a disk failure prediction method, system, terminal and storage medium. Background technique [0002] The disk is the core component of the storage system. Whether the disk can provide stable and reliable data access directly affects the reliability of the entire storage system. Disks are different from solid-state drives in that they are a mixture of magnetism, electricity, and mechanics, and their inherent structure makes them less reliable. Existing large-scale storage systems often contain thousands or even tens of thousands of disks, which greatly increases the probability of disk failure. Statistics show that among the 290,000 hardware failures in the Baidu data center within 4 years, hard disks accounted for 81.84%. The disk damage rate in the Microsoft data center was 3 / 10,000, that is, 3 out of 10,000 disks failed every day. According to the ...

AI Technical Summary

Problems solved by technology

[0005] Disk life prediction has always been a difficult problem in the industry. Although data of multiple SMART characteristics can be obtained, it is not easy to find the relationship between these data and disk life.

Images