Client request separation management method and device based on distributed file system

Abstract

The present application discloses a client request separation management method and device based on a distributed file system. The method comprises the following steps: after receiving at least one client request, separately storing all client requests into different software layers session, so that different sessions can separately send own stored client requests to a metadata server corresponding to the client request, thereby bringing convenience for the metadata server to process the client request. It can be seen that after receiving the client request, all the client requests are storedin the same container for management not by adopting a unified management manner; instead, all the client requests are separately stored in different sessions; thus, the respective client requests canbe sent to the metadata server corresponding to each client request through different sessions; further, the concurrence function of each session of the client is realized through such a separate management manner, thereby effectively taking the advantages of the client multicore system and improving the system performance.

Description

technical field [0001] The present application relates to the technical field of distributed storage file systems, and in particular to a client request separation management method and device based on a distributed file system. Background technique [0002] At present, with the advent of the data age, the amount of data that needs to be stored daily is increasing. In this context, the storage server of the distributed file system has begun to be widely used. [0003] At present, after receiving the client request, the storage server of the distributed file system will store all client requests in a unified container for management, for example, in a unified system process, but with the development of the big data era With the arrival of the new era, the client requests data to grow explosively. If only one container is still used to manage all client requests in a unified manner, it will lead to poor traversal performance and no concurrency capability when performing data q...

AI Technical Summary

Problems solved by technology

[0003] At present, after receiving the client request, the storage server of the distributed file system will store all client requests in a unified container for management, for example, in a unified system process, but with the development of the big data era With the arrival of the new era, the client requests data to grow explosively. If only one container is still used to manage all client requests in a unified manner, it will lead to poor traversal performance and no concurrency capability when performing data queries on containers that store resources. problem, and, since the current client systems are mostly multi-core systems, this leads to the fact that the current unified storage and management methods for client requests cannot take advantage of multi-core distributed systems

Images