Research on cluster load analysis and scheduling strategy in cloud environment

Research on cluster load analysis and scheduling strategy in cloud environment

Hangzhou should be June

This paper first studies the characteristics of the traditional data center and its shortcomings, analyzes the characteristics of the cloud data center, and then studies the cluster of load monitoring technology and based on libvirt virtual machine load collection technology. This paper introduces the concept and characteristics of dispatching in cloud environment, and analyzes the scheduling mechanism in cloud environment with OpenStack Cloud Platform as an example. Taking Cloudsim as an example, the cloud simulation technology is analyzed and the main modules are studied. Then, in order to understand the load characteristics of cloud data center, this paper collects the load data of the production cluster in the actual public cloud, involving 1082 virtual machine instances and 100 physical machines, the time span is April 11, 2013 to April 17. The load characteristics of the cloud data center are analyzed from different angles, including the virtual machine distribution characteristics and scheduling mechanism, the virtual machine memory distribution characteristics, the virtual machine CPU and I/O characteristics, the physical node load characteristics, the virtual machine and the physical node load. Through load analysis, some observational conclusions are obtained and some optimization methods are proposed. Finally, on the basis of cluster load analysis, this paper studies the characteristics of cloud environment scheduling and virtual machine scheduling model, and puts forward an optimization algorithm ERSG (energy Reducing and SLA Guarantee).

Research on cluster load analysis and scheduling strategy in cloud environment