Research On Resource Scheduling Algorithms In Cloud Computing

With the development of cloud computing,the study on resource scheduling becomes one of the hot research areas of cloud computing.Resource scheduling is the process of mapping tasks to available resources based on scheduling goals,thereby an effective mapping relationship is established.Some scholars find that the resource schedule problem is an NP-complete problem.At present,there are many studies on resource scheduling,but still some problems are unsolved,for example,some algorithms are not comprehensive enough to consider the problem,and some algorithms have insufficient performance optimization,etc.Therefore,how to design effective resource scheduling strategies to improve the scheduling efficiency and save energy-cost have become new research hotspots.In Max-Min algorithm,the long tasks are executed on the physical machine with fast CPU processing speed,and the minimum task completion time is considered as the scheduling target.However,the drawbacks of the Max-Min algorithm are load imbalanced,large makespan and long completion time.Therefore,an optimized max-min scheduling algorithm based on the max-min algorithm(OPMMA)is introduced in this thesis.There are three different scenarios in OPMMA algorithm.In OPMMA scheduling algorithm,we use different schedule strageties according to the different characteristics of the three scenarios,the OPMMA algorithm can improve the processing rate of physical machines per unit time and reduce the scheduling time.The simulation results show that the OPMMA algorithm can lead to significant performance gain in scenario 1 and scenario 2 in terms of makespan and response time.Therefore,the algorithm has certain applicability and practicability.In order to solve the high energy consumption problem in the cloud computing environment,an energy-saving scheduling algorithm based on max-min(ESSAMM)is proposed.In the ESSAMM algorithm,the energy consumption,Qo S and resource utilization are considered at the same time.In the algorithm,the set of physical resources that meet the user's time requirements will be selected firstly,and then the energy estimation model will be used to calculate the energy consumption value generated by each task on the physical resources.Therefore an optimal mapping relationship from cloud tasks to physical machines will be established.The simulation results show that the ESSAMM algorithm has longer completion time,but the ESSAMM algorithm has the largest resource utilization and the lowest energy consumption.