Research On Workflow Task Security Scheduling Algorithm Based On Resource Clustering In Cloud Environment

With the rapid development of the Internet, cloud computing becomes the focusof the industry, academia, government and others in all walks of life. Cloudcomputing is the development of the distributed computing, parallel computing, utilitycomputing and grid computing. It is a new computing model which is based onvirtualization technology, it takes the network as the carrier, and integrates large-scaledistributed computing resources for the collaborative computing. Facing with themassive cloud resources, how to allocate resources to meet the demand of cloud usersis the key problem in the study. At the same time, the cloud environment has manycharacteristics, such as openness, heterogeneity, uncertainty and so on, security factoris the critical content in the workflow task scheduling and resource allocation process.This thesis made further research of the workflow task scheduling in cloudenvironment, through analyzing the characteristics of the cloud computingenvironment, a scheduling algorithm based on resource fuzzy clustering was proposed.This algorithm improved the task prioritization method of the classical HEFTalgorithm, it combined with the fuzzy mathematics theory, and defined differentfeatures to describe the resources’characteristics, and then it clustered the resources toget the clustering with different comprehensive abilities. Therefore, the algorithmnarrowed the scope of the task to select resources, thus it reduced the makespan of theworkflow task.During the workflow task scheduling process, there might be data leaks andother risks, or some security issues caused by the resource node failure and jitter.These problems could lead to the task to be rescheduled or delayed. On the basis ofthe algorithm based on resource fuzzy clustering, this thesis proposed a novelcomputational model，this model was combined with the scheduling mechanism andsecurity mechanism. Meanwhile, a workflow task scheduling algorithm integrated of security factors was designed. This algorithm quantified the task securityrequirements in the form of security functions, determined the security service levelof resources, so that the task would choose the scheduling resource to satisfy thesecurity requirements of tasks.Compared with the traditional scheduling algorithms, the experimental resultsshow that the proposed resource fuzzy clustering based scheduling algorithm can beeffective to divide the resources, and it also reduces the length of the workflowscheduling; the proposed integration of security algorithm is better to tackle the phaseseparation problem of scheduling and security, it ensures to fit the bill of good timeperformance, and has a high security scheduling performance.