EXPLORE PUBLICATIONS BY COUNTRIES

Distributed Systems, Fault Tolerance, Byzantine Faults, Resource Efficient

Many Web services are expected to run with high degree of security and dependability. To achieve this goal, it is essential to use a Web-services compatible framework that tolerates not only crash faults, but Byzantine faults as well, due to the untrusted communication environment in which the Web services operate. In this paper, we describe the design and implementation of such a framework, called RET-WS (Resource Efficient and Trustworthy Execution -Web Service).RET-WS is designed to operate on top of the standard SOAP messaging framework for maximum interoperability with resource efficient way to execute requests in Byzantine-fault-tolerant replication that is particularly well suited for services in which request processing is resource-intensive. Previous efforts took a failure masking all-active approach of using all execution replicas to execute all requests; at least 2t + 1 execution replicas are needed to mask t Byzantine-faulty ones. We describe an asynchronous protocol that provides resource-efficient execution by combining failure masking with imperfect failure detection and checkpointing. It is implemented as a pluggable module within the Axis2 architecture, as such, it requires minimum changes to the Web applications. The core fault tolerance mechanisms used in RET-WS are based on the well-known Castro and Liskov's BFT algorithm for optimal efficiency with some modification for resource efficient way. Our performance measurements confirm that RET-WS incurs only moderate runtime overhead considering the complexity of the mechanisms.