Disk state detection method and system thereof, medium and storage device

Abstract

The invention provides a disk state detection method and a system thereof, a medium and storage equipment. The disk state detection method comprises the following steps: receiving and analyzing an access request, initiating first read-write according to the access request, forming a to-be-detected queue for data blocks in a to-be-detected state, and initiating active read-write under the condition that read-write load information of a server meets a first preset performance index, and detecting that the disk is in a sub-health state by analyzing the read-write performance index of the disk. According to the technical scheme, disk read-write is realized by dynamically and actively initiating the access request, and the actual read-write performance of the disk is perceived based on the data read-write transmission performance of the operating system, so that different types of disk states are detected according to the disk read-write performance index; the disk state detection result has accurate predictability, so that the probability of data loss is reduced.

Description

technical field [0001] The invention relates to the technical field of computer applications, in particular to a disk state detection method, system, medium and storage device. Background technique [0002] Today, with the rapid growth of computer performance, the performance of disks as storage media has undoubtedly become the key to the entire system. Although traditional mechanical disks have grown rapidly in storage capacity and can be read and written infinitely, there is no significant improvement in read and write performance. Solid state Disks have obvious advantages in terms of read and write speed performance, but are expensive and have a limited number of reads and writes. Due to the consideration of disk hardware performance and cost, storage resource pools are often built in the form of hierarchical storage in the hyper-converged infrastructure. Even if different types of disks are used for data storage, the reading and writing of data often spans two levels: th...

AI Technical Summary

Problems solved by technology

[0002] Today, with the rapid growth of computer performance, the performance of disks as storage media has undoubtedly become the key to the entire system. Although traditional mechanical disks have grown rapidly in storage capacity and can be read and written infinitely, there is no significant improvement in read and write performance. Solid state Disks have obvious advantages in terms of read and write speed performance, but are expensive and have a limited number of reads and writes. Due to the consideration of disk hardware performance and cost, storage resource pools are often built in the form of hierarchical storage in the hyper-converged infrastructure. Even if different types of disks are used for data storage, the reading and writing of data often spans the two levels of performance layer and capacity layer.

In the existing technology, in the business scenario of hierarchical storage, based on the principle differences of different types of disks in storage media and core architecture, there is a large gap in the frequency of disk data reading and writing. When the read and write operations are in progress, it is possible to find that there is a problem with the health status of the disk, resulting in the inability to predict the read and write performance of disk data, lack of predictability for the disk state, and greatly increase the probability of data loss

Images