Massive small file migration method for use in hierarchical storage system

Abstract

The invention discloses a massive small file migration scheme for use in a hierarchical storage system, belongs to the technical field of storage, and aims to solve the problem of low massive small file transmission efficiency in an existing migration scheme. The method is applied to a hierarchical storage framework provided with a front-end Lustre file system and a back-end HDFS (Hadoop Distributed File System), and comprises a file request processing step, a file correlation calculation step, a threshold calculation step, a small file combining step, a data re-deleting step and a data transmitting step. A device comprises a request processing module, a file correlation module, a threshold calculation module, a small file combining module, a data re-deleting module and a data transmitting module. Through adoption of the method, the advantages of the Lustre file system and the HDFS are brought into full play; performance bottlenecks in small file archiving and reading processes are analyzed completely; the small files are transmitted by an optimal scheme at any moment through quantitative threshold calculation; and the migration efficiency of the massive small files between two stages of file systems is increased.

Description

technical field [0001] The invention belongs to the technical field of storage, and more particularly relates to a method for migrating a large number of small files in a hierarchical storage system. Background technique [0002] With the rapid development of computer and network technology, global data is growing explosively. According to statistics, the total amount of global data in 2020 is expected to exceed 40ZB (about 4 trillion GB), which is 22 times that of 2011. In the past few years, The global data volume is growing at a rate of 58% per year. Large-scale Internet sites need to store and manage massive amounts of data, requiring the use of a large number of servers. [0003] With the growth of data and access volume, the development of distributed technology to solve big data and high concurrency problems is becoming more and more mature. In terms of computing, any high-configuration stand-alone machine can no longer meet the performance requirements, so distribu...

AI Technical Summary

Problems solved by technology

[0008] In view of the above defects or improvement needs of the prior art, the present invention provides a method for migrating a large number of small files in a hierarchical storage system, the purpose of which is to solve the technical problem of low transmission efficiency of a large number of small files existing in the existing migration scheme

Images