计算机科学 ›› 2009, Vol. 36 ›› Issue (8): 158-160.

• 软件工程与数据库技术 • 上一篇    下一篇

带有周期检测性的软件预防性再生策略

赵旭峰,钱存华,NAKAGAWA Toshio   

  1. (南京工业大学管理科学与工程学院 南京 210009);(Department of Marketing and Information System, Aichi Institute of Technology, Aichi 470-0392)
  • 出版日期:2018-11-16 发布日期:2018-11-16
  • 基金资助:
    本文受国家自然科学基金项目(70471077,70801036),教育部人文社科规划基金(05JA630027)资助。

Optimal Preventive Software Rejuvenation Policy with Periodic Testing

ZHAO Xu-feng,QIAN Cun-hua,NAKAGAWA Toshic   

  • Online:2018-11-16 Published:2018-11-16

摘要: 研究了软件老化过程及软件预防性再生策略问题。引入累积损伤模型,将由aging related bug引起的内存的消耗看作是由于冲击造成的内存损伤量,通过周期性的检测可以得到内存被逐步消耗而不被释放的量,当内存消耗量达到预防性再生值时在下一个检测点执行软件预防性再生策略。建立了单位时间费用模型并解析得到使单位时间期望费用最低的预防性再生值,最后通过模拟进一步验证了最优再生策略的条件。

关键词: 软件老化,软件再生,累积损伤,维护策略

Abstract: Software aging process and preventive software rejuvenation policy were studied in this paper. Consumption of physical memory, which would not be released, caused by aging related bug, was considered as damage suffered by shocks and could be known by periodic tests, preventive software rejuvenation was done at the next test point when consumption of physical memory reaches a preventive rejuvenation value. Expected cost rate model and optimistic preventive rejuvenation value which minimizes expected cost rate was analytically derived. Simulation experiment shows the validity of the rejuvenation conditions.

Key words: Software aging, Software rejuvenation, Cumulative damage, Maintenance policy

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!