基于接纳控制和ILP资源调度模型的BDAaaS系统架构

A BDAaaS System Architecture Based on Admission Control and ILP Resource Scheduling Model

摘要:

为了使云计算平台为大数据分析提供有效支持,提出一种大数据分析即服务(BDAaaS)的系统架构；首先,当用户向系统提交大数据分析应用(BDAA)时,通过接纳控制器评估任务的执行时间和成本并作出接纳决策；然后,通过服务等级协议(SLA)管理器根据任务的服务质量(QoS)需求制定SLA；最后,利用提出的整数线性规划(ILP)资源调度模型,以最小化执行成本为目标,在满足SLA下合理调度资源来执行任务；仿真结果表明,提出的方案能够有效降低任务执行时间,具有有效性和可行性。

Abstract:

In order to make the cloud computing platform provide effective support for large data analysis, a big data analytics as a service (BDAaaS) system architecture is proposed. First, when a user submits a big data analysis application (BDAA) to the system, the admission controller is used to evaluate the execution time and cost of the task and make an admission decision. Then, the SLA is built by the service level agreement (SLA) manager according to the Quality of Service (QoS) requirements of the tasks. Finally, the resource scheduling model based on ILP with the goal of minimize the execution cost is proposed, and used to schedule resources reasonably under satisfying the SLA. Simulation results show that the proposed scheme can effectively reduce the task execution time, which is effective and feasible.