云环境下融合安全与可用性的DAG任务调度  

DAG Task Scheduling Integrating with Security and Availability in Cloud Environment

作　　者：刘亚秋[1,2] 邵洪润 景维鹏[1,2] 

机构地区：[1]东北林业大学信息与计算机工程学院,哈尔滨150040 [2]黑龙江省林业生态大数据存储与高性能(云)计算工程技术研究中心,哈尔滨150040

出　　处：《计算机工程》2014年第12期12-18,共7页Computer Engineering

基　　金：国家自然科学基金资助项目(31370565);哈尔滨市科技创新人才研究专项基金资助项目(2013RFXXJ089)

摘　　要：在异构复杂的云环境中,任务调度时应保证资源的安全与可用性。为此,提出一种融合安全与可用性的DAG任务调度策略。改进主观逻辑信任模型,将计算出的资源信誉度作为调度的重要依据,对树云资源与DAG任务的可用性偏离度进行量化,给出调度函数并设计贪心调度算法(ISAG)。实验结果表明,在资源安全性和可用性较低的环境中,ISAG算法仍能满足用户需求,与表调度算法DLS和可用性优先调度算法Afsa相比,具有更高的任务执行成功率。For the instability of security and availability of resource in the heterogeneous and complex cloud environment, a DAG task scheduling strategy is proposed to satisfy the demand for security and availability of cloud tasks, which is integrated with both security and availability in cloud computing environment. The strategy first modifies the subjective logical trust model to calculate the reputation of cloud resource that is the basis of scheduling, and then qualifies the availability deviation of tree cloud resource and DAG task,and finally defines the schedule function and greedy scheduling algorithm integrating with security and availability,called ISAG. Simulation results prove that ISAG can efficiently meet the requirement of tasks on cloud platform of low availability and low security, and gets better performance and higher execution success rate compared with DLS and Afsa.

关 键 词：云计算 树云环境 DAG任务调度 云资源安全 云资源信誉度 可用性偏离度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]