Research On Task Scheduling And Virtual Machine Resource Optimization Allocation In Cloud Environment

With the further development of computer networks,cloud computing has become a popular and efficient computing model.It not only eliminates the cost of maintaining user hardware configurations,but also provides efficient and scalable computing power.However,the random and large-scale nature of user-submitted tasks and the heterogeneous nature of virtual resource nodes have led to task scheduling and optimal allocation of virtual machine resources as the biggest challenges facing the development of cloud computing.Despite intensive research and significant progress by numerous scholars in these two areas,the following problems still exist.Currently,most of the task scheduling algorithms in cloud environments take task completion time or execution cost as the optimization goal.Such strategies,while optimising task completion time and execution cost,lead to more severe load imbalance problems for virtual resources.In addition,the existing optimisation algorithms also lack flexibility and make it difficult to obtain the optimal task scheduling strategy.The optimal allocation of virtual machine resources is one of the effective ways to solve the high power consumption and low utilisation problems in cloud data centres,however,existing studies do not consider the potential overload of physical hosts during virtual machine resource consolidation and the impact of load,cost and energy consumption problems caused by virtual machine migration.In order to solve the task scheduling challenges in the cloud environment,this paper constructs a task scheduling model based on multi-objective optimization,which defines the process and related attributes of task scheduling in cloud environment.Based on this model,this paper proposes a multi-objective optimization task scheduling slgorithm based on improved artificial fish swarm.The multi-objective optimisation is carried out with task completion time,execution cost and load balancing as evaluation indicators during the algorithm’s optimisation search,so as to obtain the optimal task scheduling strategy.Finally,the experiments show that the multi-objective optimization task scheduling slgorithm based on improved artificial fish swarm not only has significant advantages in terms of convergence speed,but also outperforms other algorithms in terms of task completion time,load balancing and execution cost.To address the problem of optimising virtual machine resources in a cloud environment,this paper constructs a load-aware based virtual machine integration model,which adopts a hybrid consolidation mechanism based on workload prediction and can effectively prevent workload fluctuations,thereby improving the efficiency of virtual resource consolidation.Based on this model,this paper proposes an energy-efficient optimization based virtual machine integration algorithm to achieve efficient and optimal allocation of virtual machine resources.Finally,the experiments show that the energy-efficient optimization based virtual machine integration algorithm can not only improve the quality of service,reliability and resource utilisation of the data centre,but also effectively reduce the energy overhead.