Research On Scheduling Algorithm For Cloud Cryptographic Job Flow

Cloud computing technology provides people with convenient,efficient,economical and flexible on-demand services,while the security challenges it faces have become the main problems of academia and service providers.The use of cryptography technology can effectively guarantee data security,network security and computing security in the cloud environment,and it is also conducive to the centralized and standardized deployment and management of cipher applications.In order to employ the computing resource in cloud effectively and hold the ability of providing service on demand,it is crucial and necessary to design an efficient and high-quality task scheduling algorithm for cipher service system in cloud.Taking it into consideration that there are multiple service requests as well as random cross concurrency of data-dependent and non-data-dependent job flow in the cloud cipher service system,a two-level cryptographic job flow scheduling system is proposed in this paper.Firstly,through the mapping of cipher function of tasks,the correct implementation of multi-job flow cipher function is guaranteed.Then,for the cross problem of different working modes in each tasks with the same cipher function,facing the static cryptographic job flow and the dynamic online cryptographic job flow respectively,a job flow scheduling strategy based on association data localization is proposed to improve the security of tasks processing and guarantee the overall service performance of the scheduling system.The main work and innovations of this thesis are as follows:(1)Analyzing the characteristics of cryptographic job flow in cloud environment and the performance demand of scheduling system,constructing a diversified description method of cryptographic job flow and its tasks and cloud cryptographic processing nodes,designing a two-level scheduling system supporting random cross cryptographic job flow.On the basis of ensuring the correct implementation of the cipher function in the multi-user,multi-algorithm and multi-processing command job flow random cross application,combined with the corresponding scheduling strategy to search and match the best processing node for tasks.(2)By designing the job flow assignment table and its query-response and amendment methods,tasks of the one data dependent job flow are assigned to the same processing node to run.The localization of the associated data in this kind of task processing isrealized,which avoids the extra transmission and security threat caused by the interaction process of the associated data between different processing nodes,simplifies the key matching and intermediate state management.(3)Taking the localization of association data as the main scheduling goal,and by separating the data-dependent and non-data-dependent job flow,scheduling strategies based on association data localization are proposed for static cryptographic job flow and dynamic online cryptographic job flow respectively,which realize parallel processing on demand when multiple job flow share limited computing resources.Finally,the task completion time,average resource utilization and scheduling fairness of the scheduling algorithm proposed in this paper are tested and analyzed by C programming language under windows.Simulation results show that the scheduling strategy in this paper not only realizes the localization of associated data,but also improves the quality of service of the scheduling system,and shows better stability,which can meet the needs of cloud cipher service.