Research On Energy Modeling And Measurement Method Of Storage System

Much more attention has been paid to data storage service because of the advent of ‘big data' and ‘cloud storage'.The development of cloud storage technology,the increase of user and the exponential growth of data had brought heavy work pressure and huge energy consumption to the data center.The energy management problem has been a great tickler that badgers the development of a data center.Excessive energy consumption brings high power costs and leds to the increase of carbon dioxide emissions.In recent years,industry and academia make some suggestions for new types of energy management technology.Most of the energy management technology only payed attention to energy saving effect,but they do not payed attention to the storage system energy consumption modeling and measurement.On the one hand the study of energy consumption modeling and measurement can obtain more accurate energy model and real-time prediction of the energy storage system;On the other hand it can reasonably estimate growth trend of electricity costs and reduce the budget.Therefore,the research on energy modeling and measurement is becoming more and more important.In this paper,the energy consumption model of mainstream storage units are studied,we develop a statistic-based hard-disk power modeling method and a gradient-based SSD power modeling method.The two methods are based on the working principle of storage unit,they have good versatility and prediction accuracy.An analysis framework is proposed for storage space utilization and static energy consumption in this paper,it can used for measuring space and power cost in cloud storage environment with data management technologys.Combined with traditional disk and SSD energy models,a method based on energy-proportionality is proposed to measure the dynamic energy of storage system.This method can solve the problem which the traditional theory of energy efficiency can not measures the energy-saving effect in hybrid storage environment.The main topics and contributions of this paper can be summarized as follows:(1)We propose a statistic-based hard-disk power modeling method that estimates the power consumption of storage workloads.The model makes up the weakness of traditional fine-grained model and it is more accurate than the coarse-grained model.In practical applications,it does not need to record the disk internal activities,and does not need to trace complex parameter.Our power estimation results are highly accurate,which means error of 3% and the model is applicable to the synchronous IO and asynchronous IO.Moreover,our model can also be applied to various online storage systems and data center.(2)We propose a gradient-based SSD power modeling method that estimates the power consumption of storage workloads,and it effectively enhances the energy consumption prediction precision.This method is based on the hierarchical structure and working principle of NAND flash chips,and analyzed the energy consumption in the process of reading and writing.we build energy consumption gradient list with alternation and parallelism operation,and predict energy consumption of SSD.This kand of modeling method will not bring additional performance overhead to the system.the experimental results show the prediction accuracy of gradient-based modeling method have been improved significantly compared with the traditional linear model.The immergence of cloud computing service and cloud storage service provide an elastic service for user.(3)We propose an analysis framework for storage space utilization and static energy consumption.we developed an intelligent solution to the problem which the existing research only pay more attention to the CPU and the utilization ratio of network and less to utilization ratio of storage system resources.This framework analysises the utilization ratio classification of of storage system and presentes related measurement method.We propose the measure of storage space utilization and analyz effect of different technologies to static energy consumption in the storage system.Example analysis confirmed that the method can effectively analyz storage system utilization and measure static energy consumption.this offer reference to data center budget.(4)We propose a method based on energy-proportionality to measure the dynamic energy of storage system.According to statistic-based hard-disk power modeling method and gradient-based SSD power modeling method.We analyze energy features of the storage system with energy-proportiona.We find the conditions of storage system efficiency and behavior characterist through the theoretical derivation and experimental verification This method can solve the problem which the traditional theory of energy efficiency can not measures the energy-saving effect in hybrid storage environment.