arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1302.2203cs.DC
收藏

商业云服务性能评估实验设计的因子框架

A Factor Framework for Experimental Design for Performance Evaluation of Commercial Cloud Services

Zheng Li, Liam O'Brien, He Zhang, Rainbow Cai

摘要:

考虑到商业云服务的多样性,候选服务的性能评估对于服务客户(例如成本效益分析)和提供商(例如服务改进方向)都是至关重要和有益的。在评估实施之前,选择合适的因素(也称为参数或变量)在设计评估实验中起着先决条件的作用。然而,似乎缺乏系统的方法来选择因素的云服务性能评估。换句话说,在大多数现有的评价研究中,评价者随机和直观地关注实验因素。本文在对商业云服务进行分类和建模的基础上,提出了一个商业云服务性能评估实验设计的因素框架。该框架概括了人们目前在云计算领域考虑的性能评估因素的实践状态,反过来可以帮助设计用于评估云服务的新实验。

英文摘要:

Given the diversity of commercial Cloud services, performance evaluations of candidate services would be crucial and beneficial for both service customers (e.g. cost-benefit analysis) and providers (e.g. direction of service improvement). Before an evaluation implementation, the selection of suitable factors (also called parameters or variables) plays a prerequisite role in designing evaluation experiments. However, there seems a lack of systematic approaches to factor selection for Cloud services performance evaluation. In other words, evaluators randomly and intuitively concerned experimental factors in most of the existing evaluation studies. Based on our previous taxonomy and modeling work, this paper proposes a factor framework for experimental design for performance evaluation of commercial Cloud services. This framework capsules the state-of-the-practice of performance evaluation factors that people currently take into account in the Cloud Computing domain, and in turn can help facilitate designing new experiments for evaluating Cloud services.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑