Automated Testing of Web Services Based on Algebraic Specifications

Abstract

The testing of web services must be done in a completely automated manner when it takes place on-thefly due to third-party services are dynamically composed to. We present an approach that uses algebraic specification to make this possible. Test data is generated from a formal specification and then used to construct and submit service requests. Test results are then extracted and checked against the specification. All these are done automatically, as required. We present ASSAT (Algebraic Specification-Based Service Automated Testing), a prototype tool that performs these tasks and demonstrate its utility by applying it to Amazon Web Services, a real-life industrial example. • finite state machines (FSMs) and their variants and extensions, such as extended FSMs [8], Stream X-Machines [9], and protocol state machines (PSMs) [10], • labelled transition systems and process algebra, such as symbolic transition system STS [11], and • various kinds of Petri-nets, such as [12].Behaviour-based formal specifications like these can specify valid sequences of service invocations, for example, but they are weak on functional correctness. Even where test cases can be produced for individual services, it is not known how to turn these into test cases for composed services. A method is proposed for Petri nets in [12], in the context of cloud computing, but as it has not been implemented, it is unclear whether it is feasible. B. Algebraic specificationAlgebraic specification was first proposed in the 1970s as an implementation-independent specification technique for abstract data types [13], [14]. Since then, it has been extended to concurrent systems, state-based systems, software components and service-oriented systems. The theoretical 2015 IEEE Symposium on Service-Oriented System Engineering 978-1-4799-8356-8/15 $31.00

Pattern-Based Approach to Modelling and Verifying System Security

“…Another advantage of algebraic specification is that it enables automated testing of software against formal specifications [23], [24]. These test automation techniques can be applied to the implementation of system security design formally specified in algebraic specifications.…”

“…Algebraic specification is a formal method that supports specification refinement, verification and software testing. It has been widely used in the formal development of objectoriented software, software components and Web services [23]- [25]. Our empirical studies [26]- [28] demonstrated that algebraic specifications are easy to learn and to understand.…”

“…Another attractive feature of algebraic specifications is that they can be used directly in automated software testing [23], including test case generation, test execution and test oracles. Thus, they can be employed to perform security testing to check if the implementation of a security solution is correct with respect to a formal model.…”

A verified and optimized Stream X‐Machine testing method, with application to cloud service certification

“…Other recent work with the same goal of testing cloud services from abstract specifications includes an automated approach that generates useful tests from algebraic specifications but does not especially address test‐coverage issues and a high‐level method for specifying different kinds of robustness tests for a cloud platform, based on input validation, or state‐space exploration, or concurrent access stress‐testing , which does not address full test automation.…”

Evaluating the Ontological Semantic Description of Web Services Generated from Algebraic Specifications

“…Algebraic specifications are at a very high level of abstraction, independent of any implementation details, and can be used directly in automated software testing [10]- [12]. However, algebraic specifications do not directly support efficient searching of services, and neither do other formal methods.…”