Research On Energy-saving Replica Placement Strategy Of Storage-type Data Center Based On Thermal Aware

With the exponential growth of data,the scale and number of data center used to store massive copies of data has gradually expanded.However,the improper placement of data replica leads to increased energy consumption in the data center and increased carbon emissions,which not only increases the operating cost of the data center,but also causes environmental damage.Researchers have found that storage equipment energy consumption and refrigeration equipment energy consumption in storage-type data center are the main components of total energy consumption,and heat recycling is a problem that causes low cooling efficiency.Therefore,the introduction of airflow organization to improve data placement and improve Cooling efficiency,reducing the total energy consumption of storage-type data centers.Based on the characteristics of the data center airflow organization model,this paper proposes energy saving strategies for replica placement and improves the placement of replicas to reduce the total energy consumption of the data center.Firstly,this paper combines the characteristics of data center airflow organization and heat recycling to build an energy consumption model for storage data centers.Secondly,it defines the copy placement access model and the replica placement area,and proposes two energy-saving strategies based on thermal-aware replica placement TARP-I and TARP-II.Finally,Cloud Sim was used to build a storage-type data center.Single-copy online and multi-copy online experiments were conducted to verify the effectiveness of TARP-I and TARP-II.Experiments show that compared with Min Tout-G and Min Tin-G,TARPI can save about 70% and 25% of total energy consumption in single-copy online experiments;Compared with Min Tout-G and Min Tin-G,TARP-II can save about 40% and 10% of total energy consumption.In addition,this paper also analyzes and discusses the impact of the four algorithms on storage device energy consumption,refrigeration energy consumption,refrigeration equipment supply temperature,replica placement and distribution,algorithm performance,and power consumption parameters.Based on the experimental analysis of multiple factors,TARP-I and TARP-II effectively reduce the total energy consumption of storage data centers and provide theoretical support for the construction of green data centers.