Study and characterization of storage device traces for determining the presence of selfsimilar model

Abstract

In this work we pretend to characterize the traces of some peripheral storage devices or disks, with the aim of determining the presence of the selfsimilar model. We have twenty-three (23) storage devices, working with financial applications, during twelve (12) hours of system activity. Each measurement represents a read/write request to the device, shaped by the variables: arrival time of the request, identification of the device where request is done, address and size of the request, and type of the operation made (read/write).We have constructed the time series: Number of request to disk at time interval (t-1,t) , and we have observed that the time series shows characteristics belonging to a selfsimilar process with a large range dependence.