On The Model Of Data Flow In Complex Network Storage System

Many kinds of spatio-temporal complex behaviors are concealed in large scale network storage systems. The state of the whole system and storage events are discrete on space and time, moreover, characteristics such as intricate topology of network architecture, complicated service rules, and dynamical I/O requests, etc, demonstrate the complexity of network storage system more obviously. Object-based Storage (OBS) system, with some intelligence and adaptability, can be abstracted as an autonomous dynamic system where space and time are discrete. As an effective approach to study dispersed space-time dynamic system, Cellular Automata (CA) is used to analyze and evaluate the rule of data flow of OBS system in complex network.By analyzing the reasons and the characteristics of data flow in complex network storage system, two CA models oriented to complex network storage system are proposed. Firstly, from the view point of whole architecture of storage network system, a cell automata model named SNCA is presented: in this model, storage network node is regard as cell, and we use an attribute of storage object called lifetime to analyze the data flow rules in storage network to ascertain the congestion degree. Simulation results show that data object flow has global relativity with the phase transition of storage network. Secondly, from the view point of storage node of OBS system, a corresponding cell automata model named OBSDCA is also presented: in this model, data migration and replication mechanism are combined to analyze hotspot migration rule and load distribution characteristic among storage nodes. Simulation results show that data distribution in OBS system characters some self-organization.Cellular automata can be utilized to simulate the internal complex behaviors and evolvement rules of complex storage system. Furthermore, it can also provide some theory to support the complex storage system of next generation network.