面向组件的软件老化抗衰策略的研究

随着软件应用系统的复杂性不断提高,系统性能衰退的现象也日益普遍,研究表明计算机应用系统性能的衰退现象主要是由于软件的老化产生的.为了缓解软件老化问题、减少软件性能衰退造成的损失,最终实现一种可准确对系统当前老化状态的预测模型,本文提出一种面向组件的软件老化抗衰策略.该策略在以往研究的基础上,细化了软件再生粒度,将老化的预测和再生扩展到应用组件级.使用灰色预测方法实现在可利用的系统资源不足的情况下的老化状态的预测.还应用Matlab对该模型进行了模拟实验.实验结果表明该策略具有良好的可用性和有效性,延长了软件系统的运行时间,提高了用户的访问成功率,并进一步增强了系统稳定性.

Research of the Components Oriented Software Aging Regeneration Strategy

With the unceasing enhancement of the complexity of application system software,system performance declines become more and more often.We can know from the researches before that the declines of computer system most caused by the software aging.In order to ease the problem of software aging and reduce the losses of software collapses,and we want to achieve a predicate model of software aging that can forecast the current statements of computer system ultimately.This paper proposes a software aging regeneration strategy which is components oriented.This strategy thins the regeneration size,and makes the research deep into the components level.We use the Gray Model to complete the prediction of the software aging with the insufficient informations of system resources.We also make a experiment with Matlab to prove our strategy.The experimental result shows the running time of software system and the rate of users’ passes increase effectively.The usability and effectiveness of this regeneration strategy based on components have been proved.This paper provides support for fine granularity of the research in this field.