Research On Virtual Machine Placement Optimization Based On Double-fitness Genetic Algorithm

Virtual machine technology is the basis of cloud computing, data center resources are the basis for cloud computing hardware and software.Along with the development of cloud computing,data centers increasingly largely,and the management is increasingly complex.Generally Data center energy consumption is prevailing excessive,and physical resource utilization is not high.And the physical resources consists of some waste, data center itself contents of various delays, physical load is not balanced and other issues. The article focuses on the virtual machines placement under the cloud circumstance. In the course of the placement, we consider minimizing the physical the waste of resources in the data center,while minimizing data center total delay,under the each physical machine load balancing. The main research work the following aspects:(1) Taking virtual machine placement in the cloud circumstance into consideration,the article analyses the architectural of the data center, creates the data center resource utilization, the waste of the physical machine resources, load balancing and data center delay(CPU process delay and network transfer delay) mathematical modal.(2) For virtual machine placement problem(similar to packing NP- difficult problem) needs to optimize multiple objectives, designing and implementing a virtual machine placement algorithm based on improved genetic algorithms.And for the first time and get fit decreasing algorithm results,then we compare performance(time and space complexity) verification algorithm, which proves that the algorithm is superior to greed algorithm, and proves it is an effective algorithm for virtual machine placement.(3) Just like the method of solving multi-objective optimization problem, we first get Pareto optimal solution set by improved genetic algorithm, and then pick Pareto front Solution Set To solve the problem.(4) Designing a virtual machine placement simulated system,witch is based on MVC mode and take the Structs1.2 as framework. And we define the physical machine class, virtual machine class, the top virtual machine placement project built into the system.Then we design the Sub module, complete each module system platform and background databases implementation.