针对企业Java业务组件（EJB）的负载请求特点，提出了基于关联对象组的EJB负载分配算法（ROGRD）。该算法对集群环境下的EJB资源创建情况进行划分，并根据划分的集合在排队论系统模型的基础上建立了代价函数和回报函数，然后依据不同的请求类型，且在请求内容的分布，服务器的负载状况以及服务器的处理性能的衰减3个约束条件下，获得代价最小而回报最大分配方案。实验结果表明，ROGRD比基于位置的请求分配算法的响应时间减少了11%，吞吐量提高了2815%。

以“八六三”项目为背景，在探讨应用服务器技术、组件和框架技术的基础上，通过结合CORBA开放性、跨平台、跨语言的特性和EJB业务处理能力，设计并实现了一个可伸缩、健壮的、支持Web分布式应用的中间应用服务器系统平台，其功能包括：组件容器、管理控制台、事务服务、外部资源管理、名字服务等；并在实际中得到了初步应用，取得了很好的效果。与其它方式的服务相比，它有着易于开发、易于管理和易于加载服务等特点。对其理论背景、体系结构、关键技术以及其上的应用领域框架等进行了全面的研究。

The unplanned system crash in application server leads to huge maintenance cost. In order to reduce the maintenance cost and prevent the failure, the optimal rejuvenation schedule is put forward based on the degradation of application server. Further, the degradation process of application server is modeled and the long-term cost is analyzed by renew processes theory. Then the boundary condition of the optimal rejuvenation schedule is discussed. Finally, the numerical result shows that the optimal rejuvenation schedule can further reduce the maintenance cost of application server.

Software aging is an important factor that affects the software reliability. According to the characteristic of performance parameters of application sever middleware, a new model for software aging prediction based on wavelet networks is proposed. The structure and parameters of wavelet network are optimized with genetic algorithm and simulated annealing algorithm. The objective is to observe and model the existing resource usage time series of application server middleware to predict accurately future unknown resource usage value. Judging by the model, we can get the aging threshold before application server fails and rejuvenate the application server before systematic parameter value reaches the threshold. The experiments are carried out to validate the efficiency of the proposed model, and show that the aging prediction model based on wavelet network with hybrid genetic algorithm is superior to the BP neural network model and wavelet network model in the aspects of convergence rate and prediction precision.

According to the characteristics of the operational behavior and runtime state of application sever, the resource consumption time series are observed and modeled by fuzzy wavelet network (FWN) with fuzzy logic inference and learning capability. The objective is to model the extracted data series of systematic performance parameters to predict software aging in application server. The dimensionality of input variables of FWN is reduced by principal components analysis (PCA), and the structure and parameters of FWN are optimized with adaptive genetic algorithm (GA). Judging by the model, we can get the aging threshold before application server failed and preventively maintenance the application server before systematic parameter value reaches the threshold. The experiments are carried out to validate the efficiency of the proposed model and show that the aging prediction model based on FWN with adaptive genetic algorithm is superior to the neural network (NN) model and wavelet network (WN) model in the aspects of convergence rate and prediction precision.