Research And Analysis For Resource Scheduling Strategy In Cloud Computing

Since the concept of cloud computing has been proposed,it quickly lead to a global research and development boom,many emerging cloud computing products are pushed to the market.Cloud computing is a computing model after distributed computing,parallel computing,grid computing,utility computing and virtualization computing,providing services according to users' requirements.Cloud computing uses virtualization technology to abstract all kinds of hardware resources of data center into virtual units,and provide various services according to users' requirements.As an important carrier of cloud computing,the cloud data center is in high-speed expansion.To meet the increasing requirements of users,the cloud resource providers and other enterprises must consider some problems about energy consumption,resource utilization,system performance,service quality,load and cost control in data center with the gradually increasing IT devices.At present,there are several main directions in the research of cloud computing resource scheduling: resource scheduling for improving the system performance,resource scheduling for improving the service quality,resource scheduling based on the economic principles,resource scheduling for improving the resource utilization rate,resource scheduling for reducing energy consumption,resource scheduling based on load balancing.There are many kinds of resources in the cloud system,such as CPU,memory,hard disk,network bandwidth,I/O equipment,etc.,and the resources in the system are changed dynamically.Users submit tasks to the cloud,the number of tasks submitted by each user is different,each task requires different resource type,and the amount of each type of resource is different,too.In the cloud computing system,cloud users and cloud providers benefit from different angles,a good resource scheduling strategy needs to balance the interests of all participants.Fair allocation strategy based on preference FABP(Fair allocation strategy based on preference)referred in this paper,gives the definition of user priority and task priority,and uses the stable matching theory to solve the problem about virtual machine placement.During the process of task scheduling,firstly,this algorithm use the user's priority to locate the user with the highest weight,then select the user's corresponding task with the highest weight according to the task priority.In the aspect of resource allocation,FABP algorithm establishes the preference relation between tasks and physical machines firstly,and then places tasks to the most appropriate physical machine through the stable matching theory,to reduce the resource fragments.The experimental results show that the algorithm can not only shorten the average task scheduling time,but also ensure the fairness of users and tasks for improving the user satisfaction and maximize the utilization of comprehensive resources to ensure the benefit of cloud providers.