混合云环境中数据敏感工作流调度

混合云环境下调度包含敏感数据的工作流主要考虑在满足数据安全性以及工作流截止时间的前提下,对工作流任务在混合云上进行分配,实现计算资源与任务的映射,并优化调度费用。采用了整数规划来建模求解包含数据敏感性、截止时间和调度费用3种约束条件的混合云工作流调度问题,同时为优化模型求解速度,基于“帕雷托最优”原理对工作流任务在混合云上的分配方案进行筛选以减小模型求解规模。实验表明,优先排除不合理的任务分配方案可有效减小整数规划模型的求解规模,缩短模型计算时间,在产生较小误差的情况下获得较优的调度结果。

Scheduling Data Sensitive Workflow in Hybrid Cloud

Using public resources to extend the capacity of private cloud is an effective way for the enterprises to achieve high efficiency and elasticity in data storage and computing.Scheduling workflow with sensitive data in hybrid cloud needs to satisfy the requirements of data security and execution deadline.In order to minimize the monetary cost,the scheduler must decide which tasks should be run on the public cloud and on which computing resource each workflow task should be allocated.Integer linear program(ILP) was used to formulate workflow scheduling problem with three objectives,such as data sensibility,deadline and cost.For the purpose of reducing the solve time of ILP model,the task assignment filter strategy based on Pareo optimality theory was designed.The filter strategy can decrease the scale of task assignments,and reduce the mappings between tasks and resources of ILP model.Experiments show that removing the unreasonable task assignment before resource allocation can decrease the ILP model scale and reduce scheduling running time,while the method can obtain a good solution.