Distributed storage file recovery method and device

Abstract

The invention provides a distributed storage file recovery method, which is applied to a hyper-fusion system, and comprises the following steps: when detecting that at least one target file is deleted, creating a multi-level chain-based blindage file; the blindage file is a sparse file generated by performing zero-write filling on a storage cell where each data block of the target file is located; storing a deletion log of the target file; the deletion log comprises a file path and position information of a target file; the file path comprises a directory where the target file is located and a file name; the position information comprises a unique number of a storage unit where each data block of the target file is located, the number of layers where the built shelter file is located and a unique number of a storage cell where the built shelter file is located; and recovering the target file according to the deletion log. By adopting the method, the problem that the computer performance is reduced due to independent detection of massive files can be avoided.

Description

technical field [0001] The present disclosure relates to the technical field of cloud computing, and in particular to a recovery method, device, electronic equipment and readable storage medium for distributed storage files. Background technique [0002] Ransomware is a common and extremely harmful new type of computer virus, which is mainly spread through the Internet. The workflow of the ransomware virus is: first find the file to be attacked, obtain the file and related information to generate a public key and private key, then read the file data, encrypt it to generate a new file, and delete the original file. [0003] In related technologies, by monitoring the I / O (Input / Output, input / output) operation of each file in the disk, and then deploying the stained file, the I / O operation of the stained file is monitored. When the operation state is found to be abnormal through monitoring, obtain For all current processes of the terminal, if a certain process satisfies any sp...

AI Technical Summary

Problems solved by technology

CDP includes timing and real-time modes, but timing CDP will lose data updates in the period, while real-time CDP needs to capture and copy each I / O. For files that are more frequently modified, such as the underlying files of the database, real-time CDP technology consumes a lot of system resources and affects the performance of the database

At the same time, as the capacity of the disk becomes larger and more files can be stored, when monitoring the modification of files on the disk, it is necessary to traverse all the files on the disk, occupying a large amount of disk bandwidth, and calculating the file verification value also requires a large amount of CPU (central processing unit, central processing unit) resources

[0004] To sum up, in the unified protection of a large number of files in the disk, how to avoid the degradation of computer performance caused by the separate detection of massive files is an urgent problem to be solved at present

Images