File data accessing method, apparatus and system

Abstract

The invention provides a method, a device and a system for file data access, and the file data accessing method comprises the following steps: a file metadata request message is sent to a metadata server; file metadata returned by the metadata server is received; and according to the comparison result of the file metadata, a preset file size value and preset hotspot frequency, all the file data is obtained from a local cache of the client terminal or a high-speed memory of a data server. The file data accessing method, the device and the system realize multilevel caching of files with small hotspot, and can obtain all the file data from the local cache or the high-speed cache of the data server especially for the files with small hotspot, thus improving the reading efficiency of the files with small hotspot in network memory, and further raising the access speed of the files with small hotspot and reducing network flow.

Description

technical field [0001] The invention relates to network storage technology, in particular to a file data access method, file metadata sending method, device and system. Background technique [0002] With the rapid development of computer networks and information technology, the global information storage capacity has increased rapidly. At the same time, network storage technology has continued to develop, distributed storage has been widely used, the network scale has continued to expand, and there are more and more access nodes. In many applications, such as WEB 2.0, the amount of data is huge, but usually only a small part of the content is accessed frequently, and a large number of clients access a storage node, and the file reading efficiency is very low, resulting in uneven network load and increased access delay And other issues. [0003] In the process of realizing the present invention, the inventors found that there are at least the following problems in the prior ...

AI Technical Summary

Problems solved by technology

When the client reads the file, no matter what the file is, it will not read the entire file into the memory for the first time, and needs to send multiple I / O requests to the common storage device; and after reading, it will be cached in the current thread's In memory, cannot be shared with other threads

Different threads of the same client may also re-initiate I / O requests to read the same file, resulting in increased network interaction and disk reading, and resulting in slow file access, especially low reading efficiency for hot and small files

Images