Disk recovery method and device

Abstract

The invention aims to provide a disk recovery method and device. According to the application, the external access capability of a disk is provided through a standby data channel, each server accessesthe corresponding disk under the condition that the servers are normal, and a data access service is provided; if one server breaks down, according to the method, the access capability of the disk corresponding to the faulty disk server can be exposed to other disk servers without depending on the faulty disk server, the other disk servers take over the disk corresponding to the faulty server and continue to provide services, and on the premise that reliability and usability are guaranteed, data migration in the disk can be completed in the background.

Description

technical field [0001] The invention relates to the field of computers, in particular to a disk recovery method and equipment. Background technique [0002] Current data storage solutions include the GFS (google File System) distributed file system used in scenarios such as cloud computing and big data computing. Most of them use the method of storing multiple copies to improve data reliability. For example, the commonly used solution in the industry is One copy of data is stored in three copies, and each copy is saved to a different disk to ensure data reliability and availability through data replication. Each disk corresponds to a disk server to provide external read and write services. [0003] However, for a GFS-like distributed file system solution, when any one of the disks is inaccessible due to a disk failure or the failure of the disk server corresponding to the disk, other disks and their corresponding disk servers can still provide data access services externally...

AI Technical Summary

Problems solved by technology

[0004] However, the actual situation is far more complicated than the ideal situation. If multiple disks and their corresponding disk servers are down, all three copies of data may be unavailable, and the service can only be continued after the downtime is restored, thus prolonging the service life. failure recovery time

In addition, currently, in order to reduce storage costs, many cloud computing service providers use high-density disks. A single high-density disk is equipped with tens or hundreds of large-capacity disk units. If the disk server corresponding to a disk fails, although The disk is not damaged, but because the disk server supporting the disk read and write service fails, not only the service on the failed disk server must be migrated to other disk servers, but also the data on the disk needs to be migrated to other servers corresponding to other servers. On the disk, however, with the large volume of data replication on the disk, the data reliability recovery time may need to reach the level of hours or even days

Images