Method, device and system for reading and writing magnetic disk file of distributed file system

Abstract

The embodiment of the invention provides a method, a device and a system for reading and writing a magnetic disk file of a distributed file system, wherein the system comprises a metadata processing module cluster, and the metadata processing module cluster comprises at least two metadata processing modules; a write-in method comprises that the metadata processing module cluster receives a file write-in request from a client side, distributes all the magnetic disk files to be written-in to each metadata processing module, and each metadata processing module respectively sends the distributed magnetic disk file to be written-in to a magnetic disk; and a read-out method comprises that the metadata processing module cluster receives a file read-out request from the client side, distributes all the magnetic disk files to be read-out to each metadata processing module, and each metadata processing module respectively sends the distributed magnetic disk file to be read-out to the client side, so that data processing stress is balanced to each metadata processing module, the data processing stress of each metadata processing module is reduced, and speed and reliability for the client side in the distributed file system to write-in and / or read-out the magnetic disk file are increased.

Description

technical field [0001] The invention relates to the technical field of distributed file system data transmission, in particular to a distributed file system disk file reading and writing method, device and system. Background technique [0002] In a distributed file system, the metadata processing module is usually used to process the file that the client needs to write to the disk and / or the attribute data of the file that the client needs to read from the disk, and the file that the client needs to write to the disk And / or after the attribute processing of the file that the client needs to read from the disk is completed, the file that the client needs to write to the disk is transferred to the disk and / or the file that the client needs to read from the disk is transferred to the client. [0003] At present, in the existing distributed file system, usually only one metadata processing module is set, and only one metadata process of the metadata processing module is used to ...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the embodiments of the present invention provide a method, device and system for reading and writing disk files in a distributed file system to solve the problem that in the prior art, only one metadata processing module is usually provided, and only one element of the metadata processing module The data process processes and transmits the attribute data of the file that the client needs to write to the disk and / or the attribute data of the file that the client needs to read from the disk, resulting in extremely high data processing pressure and high data processing load for the metadata processing module , reducing the rate at which the client writes and / or reads disk files in the distributed file system, and at the same time, also reduces the reliability of the client writing and / or reading disk files in the distributed file system

Images