分布式计算环境中的协同分配任务调度仿真系统

摘要

协同分配是在分布式计算环境中进行资源分配的一种重要技术,用于把一个应用程序分解为多个子作业,然后将其分配到多个资源上同时处理来满足特定的性能要求.本文提出了一个离散事件驱动的网格资源协同分配仿真系统,实现了对用户、调度器、协同分配器、协同预留器等协同分配相关实体的仿真,实现了FCFS、FPFS、Backfill等主要的协同分配调度算法和策略,可用于资源协同分配相关的分布式计算环境的资源管理和调度算法的仿真和研究.%Resource co-allocation is one of the crucial technologies affecting the utility and quality of services of large-scale Grid-like distributed environments by simultaneously allocating the requested multiple resources to a single submitted application to meet the specific performance requirements. This paper presents a discrete-event driven simulator for studing the co-allocation-related resource management and scheduling algorithms. This tool models the main Grid components and their functions for co-allocating grid jobs, including Grid user,metascheduler, co-allocator, and co-reservator. Some main scheduling algorithms and policies, including First Come First Served, First Processor First Served and Backfilling, are implemented.