Scholarlink

<p>Response time is a vital performance metric of a big data system, indicating the system’s efficiency and measuring users’ satisfaction in today’s data-intensive environments. Through a case study, this paper presents a general framework for modeling and evaluating the response time of a Spark system using stochastic Petri nets (SPNs). To deal with the state-space explosion problem encountered in such discrete-event modeling of large-scale systems, stochastic simulation algorithms have been proposed. The response time can therefore be automatically derived to quantitatively evaluate the system. Experimental results demonstrate the impact of the number of applications and the executing task rate on this performance metric. Consequently, the framework provides a practical methodology for performance prediction, whose deployment for analyzing complex architectures benefits from high-performance computing (HPC) resources.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Response-time analysis of a big data system with stochastic petri nets

  • 期刊论文

  • 发表日期：

    2026年4月13日

查看全文
  • Xiang Wei,
  • Jiancheng Sun,
  • Jie Ding,
  • Xu Zhao,
  • Feng Zhu,
  • Yunyue Xie

摘要

Response time is a vital performance metric of a big data system, indicating the system’s efficiency and measuring users’ satisfaction in today’s data-intensive environments. Through a case study, this paper presents a general framework for modeling and evaluating the response time of a Spark system using stochastic Petri nets (SPNs). To deal with the state-space explosion problem encountered in such discrete-event modeling of large-scale systems, stochastic simulation algorithms have been proposed. The response time can therefore be automatically derived to quantitatively evaluate the system. Experimental results demonstrate the impact of the number of applications and the executing task rate on this performance metric. Consequently, the framework provides a practical methodology for performance prediction, whose deployment for analyzing complex architectures benefits from high-performance computing (HPC) resources.