File access method, data server and file access system

Abstract

An embodiment of the present invention provides a file access method, a data server, and a file access system, wherein the method includes: receiving an access request sent by a terminal; if the access operation type is a write operation, and the access request also includes data to be written, then Obtain the first data corresponding to the storage address at the first moment, generate the signature of the first data according to the first data, and modify the first data to the data to be written; obtain the second data corresponding to the storage address at the second moment, and Generate the signature of the second data from the second data; judge whether the signature of the first data is the same as the signature of the second data; if they are the same, write the modified first data into the storage location corresponding to the storage address, and generate the first access operation result , and send the first access operation result to the terminal. The file access method provided by the embodiment of the present invention can realize concurrent file access without introducing deadlock problem.

Description

technical field [0001] The invention relates to the technical field of file management, in particular to a file access method, a data server and a file access system. Background technique [0002] With the rise of the wave of big data, the importance of the Distributed File System (DFS for short) is becoming more and more obvious. DFS is based on the client / server model, which forms a file system network with many nodes. Each node can be distributed in different locations, and the communication and data transmission between nodes are carried out through the network. When people use DFS, they don’t need to care about the data source. From which node, the data in DFS can be managed like a local file system. As a large-capacity carrier, DFS's concurrent access becomes the key to the file system. [0003] At present, in order to solve the problem of high concurrent access, the lock mechanism is usually used. The so-called lock mechanism refers to constructing an exclusive key ...

AI Technical Summary

Problems solved by technology

Processes that have not obtained key resources (locks) usually check the lock status by polling, which has a large performance loss

[0004] However, the lock mechanism is prone to deadlock problems. The so-called deadlock refers to the blocking phenomenon caused by competition for resources or communication between two or more processes during execution. Unable to move forward, at this time, the system is in a deadlock state, and the file access operation cannot be completed

Images