计算机科学与探索 ›› 2011, Vol. 5 ›› Issue (3): 256-266.

• 学术研究 • 上一篇    下一篇

应用服务器资源动态重配置再生方法

刘宏磊, 杜小智, 侯 迪, 齐 勇   

  1. 西安交通大学 电子与信息工程学院, 西安 710049
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-03-01 发布日期:2011-03-01

Application Server Rejuvenation Based on Resources Dynamic Reconfiguration

LIU Honglei, DU Xiaozhi, HOU Di, QI Yong

  

  1. School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an 710049, China

  • Received:1900-01-01 Revised:1900-01-01 Online:2011-03-01 Published:2011-03-01

摘要:

传统的基于模型的老化与再生研究方法已不适于动态开放的网络环境。以对服务器内部状态、行为、环境等信息进行运行时监测, 并准确、全面地评价老化趋势为基础, 采用资源动态重配置方法提高系统服务能力、灵活性与适应性具有重要意义。以J2EE应用服务器为实验对象, 对基于资源动态重配置的应用服务器软件再生方法进行研究与实现。给出动态重配置再生模型, 提出运行时监测体系结构, 从用户体验和服务器运行状况角度给出软件老化指标、服务器端监测模型、客户端监测模型, 实现了负载发生器、用户体验数据监测等功能模块。研究资源动态重配置过程, 分析可动态重配置的资源对象, 提出将分析现有数据与预测未来数据相结合, 再辅以人工控制的决策算法。建立测试系统, 对再生模型进行实验, 验证了老化现象与再生方法的有效性。

关键词: 软件老化, 软件再生, 运行时监测, 资源动态重配置, 应用服务器

Abstract: The traditional model-based research of aging and rejuvenation hasn’t applied to dynamic open network environment. It has important significance to monitor the real-time internal state, behavior and environment infor-mation of server by run-time, make accurate and comprehensive evaluation of aging, and use resources dynamic reconfiguration to improve the system service capacity, flexibility and adaptability. This paper studies and imple-ments the software rejuvenation for J2EE application server based on resources dynamic reconfiguration. It proposes a software rejuvenation model based on resources dynamic reconfiguration, and a run-time monitoring structure, gives the index showing the aging from user experience and server status, server monitoring, client monitoring including load running, monitoring of user experience and other modules. It studies the process of resources dynamic reconfiguration, discusses reconfigurable resources, and proposes an algorithm by analysis of existing data and prediction of future data supplemented by artificial control. Experiments are shown to validate the effectiveness.

Key words: software aging, software rejuvenation, run-time monitoring, resources dynamic reconfiguration, application server