中国科学院软件研究所机构知识库
Advanced  
ISCAS OpenIR  > 中科院软件所  > 中科院软件所
题名:
WEB应用服务器容错相关技术研究
作者: 左林
答辩日期: 2007-06-06
授予单位: 中国科学院软件研究所
授予地点: 软件研究所
学位: 博士
关键词: Web应用服务器 ; 中间件 ; 容错 ; 可靠性 ; 复制 ; 适应性
其他题名: Research on Fault-Tolerant Technologies in Web Application Server
摘要: 随着Internet迅速发展,网络软件平台已成为当今软件技术发展主流趋势和竞争焦点,其核心和关键是位于中间层的Web应用服务器。Web应用服务器是在Web计算环境下产生的一种新型中间件,为创建、部署、运行、集成和管理Web应用提供一系列运行时基础服务。随着Web应用服务器在关键任务领域应用程度的不断加深,应用系统对可靠性提出了越来越高的要求,迫切地需要Web应用服务器能够提供强有力的容错支持。 Web应用服务器的容错技术研究虽然已有大量的研究工作并取得一定的成果,但是面对新的Web计算环境,仍然存在许多不足:缺乏对多层架构下事务性应用的支持,无法为用户提供端到端的可靠性保证,因而不能满足新型应用的实际需求,如电子商务、B2B应用等;对应用的QoS需求支持不足,面对应用需求的多样性和多变性,不能适应性地为应用提供有效的QoS保障。为了有效的解决这些问题,本文将容错技术与可靠性保障的其它方面如事务处理、QoS约束等技术相结合,以求对上层应用提供更强和更加灵活的容错支持。 首先本文针对Web应用服务器的多层体系结构,在已有工作的基础上提出了适用于Web应用服务器环境下的端到端可靠性定义,它同时保证用户请求和所关联事务的恰好一次(Exactly-Once)执行。我们提出了一种能够保证端到端可靠性的容错机制,该机制将复制技术和事务处理技术结合起来,一方面利用了复制技术的存活特性来保护中间层的应用操作;一方面利用了事务处理技术的安全特性来保护后端的应用数据。同时,通过交互协议来保证两者的正确协作,从而为上层应用提供了更强的可靠性保证。 在Web计算环境下,系统的多样性和复杂性导致可能出现多种复杂的事务模式。当前的容错技术往往是基于简单的事务模式,无法为实际应用中存在的复杂事务模式提供足够的支持。本文基于事务的调用方式识别出了Web应用服务器环境下几种典型的事务模式。针对这些模式,本文提出了状态同步点的概念,包含两种状态同步点:操作型同步点主要针对于客户端发起的事务调用,在事务边界内执行状态同步;事务型同步点主要针对中间层容器发起的事务调用,状态同步发生在事务边界之外。基于同步点概念,我们提出了一种支持复杂事务模式的容错机制,该机制不仅能够提供端到端可靠性保证,而且能够忍受事务协调者的失效,避免了分布式事务两阶段提交过程中的阻塞问题。 组件复制是Web应用服务器实现容错的基本技术,一些关键任务除了需要满足高可靠外,往往还需要满足一定的服务质量保证,如响应时间约束和一致性约束。这就要求在动态变化的QoS需求和系统运行环境下,容错服务设施能够合理进行组件副本的选择。本文提出了基于域的组件副本选择模型,将组件副本被划分为强一致性域和弱一致性域。域间利用一致性窗口机制进行更新传播,通过窗口大小的调整来控制副本间的不一致性程度。基于该模型,本文提出了一种基于分区加权的自适应组件副本选择算法,该算法利用动态反馈的信息来执行副本选择。为了适应请求一致性约束的动态变化,本文提出了一种一致性窗口重配算法,该算法能够根据当前一致性约束需求的变化,动态地对一致性窗口进行适应性的调整。 最后,本文给出了一个Web应用服务器的容错支撑系统WAS-FT,它对上述研究成果进行了原型实现和验证。WAS-FT构建在网驰平台的Web应用服务器OnceAS之上,容错功能以服务的形式通过标准接口与OnceAS进行集成,具有很好的灵活性和可扩展性。
英文摘要: With the explosive development of Internet, software platforms for Web applications have been the mainstream category in software industry. The core and key part of the platform is Web Application Server (WAS), which resides at the middle tier in the three-tier client/server architecture. WAS solves the problems in applying traditional middleware to Web computing environment and provides a series of run-time services to support for the deployment, integration, and execution of Web applications. As the WAS is widely applied in more and more critical industries, it is urgent for the WAS to provide high reliability guarantee for business systems. Fault-tolerance technology for Web application server has been widely studied and made great progress. However, current research still has some limitations. On the one hand, fault-tolerant support for multi-tier applications is inadequate, especially for transactional applications, and cannot provide end-to-end reliability which is important and necessary for applications, such as e-commerce and B2B applications; On the other hand, support for application’s QoS is not enough and cannot adaptively provide effective guarantee for QoS requirements of different applications. To solve these issues, this thesis combines fault-tolerance technology with some other reliability guarantee technologies, such as transaction processing and QoS, to provide stronger and more flexible reliability support. Firstly, we proposed an end-to-end reliability definition adapted to the multi-tier architecture of WAS on the basis of existing work, which simultaneously guarantees the exactly-once execution semantics of a request and the consistency of application states. Then, we present a fault-tolerant scheme for Web application server named J2EE-FT, which combines replication with transaction processing, and can provide stronger end-to-end reliability guarantee for applications. This scheme utilizes the liveness property of replication to protect middle-tier application operations, and utilizes the safety property of transaction processing to protect application data in back-end databases. In addition, we prove the end-to-end reliability property of this scheme on the basis of the definition of end-to-end reliability. Because of the dynamics and diversity of applications in Web computing environment, many complex transaction patterns may occur. Current research works only assume simple transaction pattern, and lack enough support for some complex patterns in practice. We first recognized four typical transaction patterns in Web application server. Then, we present a concept of State Synchronization Point (SSP), which includes two types SSP: Transactional Synchronization Point (TSP) and Operational Synchronization Point (OSP). TSP is applicable to the case of client transaction and state synchronization takes place out of the boundary of a transaction, and OSP is applicable to the case of container transaction and state synchronization happens in the boundary of a transaction. We propose a fault-tolerant scheme based on SSP to uniformly provide reliability support for different transaction patterns. This scheme not only can realize end-to-end reliability guarantee but also can tolerate the failure of transaction coordinator and avoids the blocking problem of 2PC protocol during distributed transactions processing. Component replication is a basic technology to implement fault-tolerance in WAS. However, the processing of user request needs satisfy some specific QoS constrains such as response time or consistency besides of high reliability. Therefore, the fault-tolerant service provided by WAS must effectively select component replica according to dynamic changes of QoS requirements and runtime environments. We proposes a domain based adaptive replica selecting model named DARSM, in which component replicas are organized into strong consistency domain and weak consistency domain, and a scheme based on consistency window is utilized to synchronize states between theses domains, and the degree of inconsistency is controlled by changing the size of consistency window. Based this model, we present a partition-balanced based adaptive replica selection algorithm PWARS, which utilizes dynamic performance metric information to select the appropriate replica set that satisfies specific QoS constraints. To adapt to the dynamic change of consistency constraints, a consistency window adaptive reconfiguration algorithm CWAR is presented, which can adaptively adjust the configuration of consistency window according to current consistency constraints distribution. Finally, we implemented a prototype of the fault-tolerant system, named WAS-FT, and applied the approaches and solutions proposed in this thesis. The WAS-FT was built on the J2EE application server, OnceAS, which is a core middleware of ONCE platform. The fault-tolerant functions of WAS-FT are implemented as system services and integrated with OnceAS through standard interface, and has good flexibility and extensibility.
语种: 中文
内容类型: 学位论文
URI标识: http://ir.iscas.ac.cn/handle/311060/6794
Appears in Collections:中科院软件所

Files in This Item:
File Name/ File Size Content Type Version Access License
10001_200318015001006左林_paper.pdf(3207KB)----限制开放-- 联系获取全文

Recommended Citation:
左林. WEB应用服务器容错相关技术研究[D]. 软件研究所. 中国科学院软件研究所. 2007-06-06.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[左林]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[左林]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2017  中国科学院软件研究所 - Feedback
Powered by CSpace