利用神经网络识别存储系统I/O特性

I/O Characteristics Identification of Storage System by Neural Network

大数据推动了存储数据呈现爆炸性增长,导致存储系统的性能与能耗问题日益突出,因此需要对存储系统进行性能与节能优化。在存储系统优化中,存储系统的I/O特性识别是优化的基本依据。为此,本文提出一种基于神经网络的存储系统I/O特性识别方法,利用神经网络技术识别存储系统I/O请求的顺序性与随机性,以及存储系统的实时性能需求等特性。实验的结果表明,该方法能够高效准确识别存储系统的I/O特性。

Big data has driven explosive growth in stored data,leading to increasingly prominent performance and energy consumption issues in storage systems.Therefore,it is necessary to optimize the performance and energy efficiency of storage systems.In storage system optimization,identifying the I/O characteristics of the storage system is the fundamental basis for optimization.To this end,this article proposes a neural network-based method for identifying the I/O characteristics of storage systems,utilizing neural network technology to identify the sequence and randomness of storage system I/O requests,as well as the real-time performance requirements of the storage system.The experimental results show that this method can efficiently and accurately identify the I/O characteristics of the storage system.