CAJ | 학술논문

本文探讨了MapReduce框架，它为控制成本、优化集群资源利用提供了保证。对于分布式计算层面，资源优化有两个途径：一是通过精细化的资源调度来保证全局资源的最大化利用；二是提升原有计算作业的资源使用率。HCE计算框架主要关注后者。
본문탐토료MapReduce광가，타위공제성본、우화집군자원이용제공료보증。대우분포식계산층면，자원우화유량개도경：일시통과정세화적자원조도래보증전국자원적최대화이용；이시제승원유계산작업적자원사용솔。HCE계산광가주요관주후자。
This paper discusses the Map Reduce framework,it provides a guarantee on cost control,and optimization of cluster resource.For the resource optimization of the distributed computing,there are two ways：one is through the refinement of resource scheduling to ensure the global maximum utilization of resources;the other is to upgrade the utilization of original computing resource.HCE computational framework is primarily concerned with the latter.