| With the increasing development and widespread application of Web Services technology and distributed system, the reliability, availability and security have become increasingly demanding. The high-available systems must provide correct services without interruptions in some parts failure or under attacks. Byzantine faults such as software bugs, operator mistakes, and malicious attacks are the major cause of service interruptions. A Byzantine fault is an arbitrary fault that occurs during the execution of an algorithm by a distributed system. It encompasses those faults that are commonly referred to as crash failures and send-omission failures. When a Byzantine failure has occurred, the system may respond in any unpredictable way, unless it is designed to have Byzantine fault tolerance. Distributed web services providers are often required that they can continue to provide services in non-secure or non-reliable network environment. Web services architecture defend against not only the benign faults but also can tolerate Byzantine faults.In this paper, the main work and innovations are as follows:This paper first introduced the fault tolerance research, and summed up a variety of solutions in synchronous and asynchronous environments, and analyzed the use of state machine replication technology and quorum certificate techniques to build fault-tolerant system, and researched the process of Byzantine fault tolerant algorithm deeply.The original Byzantine fault tolerant algorithm relies on public-key cryptography to sign messages, which is the main performance bottleneck. This paper introduced the message authentication code to substitute public-key signature, and proposed a new concept of vector of message authentication codes. To solve the problem that a receiver can not prove the authenticate of the message to a third party, this paper modified and improved the view change protocol by integrating certificate technology and message acknowledgment technology, letting nodes ask for the certificates voluntarily, so that we designed a new algorithm called BYZMV. Most of the researchers in the past establish themselves own protocol to achieve consistency between nodes. This paper analyzed the solution of how to set up the Byzantine fault tolerant web services system on Internet, which will package fault-tolerant protocol into common SOAP to avoid its own protocol being filtered by the firewall. This paper implemented a fault tolerant Web Services which is transparent to the user.This paper used Axis2C as a Web Services engine, and implemented a fault tolerant web services client called FTWSC, and a server side called FTWS. At the same time we designed a fault-tolerant framework and implemented several key components of FT module: message processors of out-flow and in-flow, which shows the feasibility of Byzantine fault-tolerant Web Services architecture. |
