Ali Aït-Bachir

Diagnosing Incompatibilities in Web Service Interactions for Automatic Generation of Adapters

Interactions between two applications encapsulated into Web services consist in series of message exchanges that must conform to service interfaces. The study reported in this text aims at dealing with the issues that arise when interactions between two services (a client and a provider) fail because their interfaces are incompatible. This may happen because the provider has evolved and its interface has been modified. It may also happen because the client decided to change for another provider which addresses the same needs but offers a different interface. The contribution of the proposed approach is twofold. First, given two services, all incompatibilities that may exist between their interfaces are automatically detected and classified into patterns. Second, according to the patterns that have been recognized, an adapter is then automatically generated. This latter is intended to act as an intermediate between the client and the provider, therefore seamlessly reconciling interactions between them.

BESERIAL: Behavioural Service Interface Analyser

In a service-oriented architecture, software services interact by means of message exchanges that follow certain patterns documented in the form of behavioural interfaces. As any software artifact, a service interface evolves over time. When this happens, incompatibility problems may arise. We demonstrate a tool, namely BESERIAL, that can pinpoint incompatibilities between behavioural interfaces.

Diagnosing and Measuring Incompatibilities between Pairs of Services

This paper presents a technique which detects all behavioural incompatibilities between two service interfaces (a client and a provider). This may happen because the provider has evolved and its interface has been modified. It may also happen because the client decided to change for another provider which addresses the same needs but offers a different interface. Unlike prior work, the proposed solution does not simply check whether two services are incompatible or not, it rather provides detailed diagnosis, including the incompatibilities and for each one the location in the service interfaces where these incompatibilities occur. A measure of similarity between interfaces which considers outputs from the detection algorithm is proposed too.

Compatibility and Reparation of Web Service Processes

When two Web services work together, they exchange messages in a predefined interface process. Two interface processes should be compatible when they can work properly. Our idea to fix incompatibility problem in service processes is to change an incompatible process so that the new process can simulate a compatible process. We consider not only the control flow but also the data flow in modeling the processes into FSMs. We present a technique that not only detects the incompatibility, but also provides resolution strategies to generate the new process.

Reconciling Web Service Failing Interactions. Towards an Approach Based on Automatic Generation of Mediators

Interactions between Web services are based on interfaces which describe Web services on both structural and behavioural perspectives. It can happen that the interface provided by a service does no longer match (for instance, because of an evolution) the interface required by its partners. In this situation, and until the required interfaces are fixed, interactions cannot succeed. To address this issue, and focusing on the behavioural part of interfaces, we propose an approach based on a mediator, automatically generated, which aims to seamlessly resolve incompatibilities during service interactions.

Detecting Behavioural Incompatibilities between Pairs of Services

We present a technique to analyse successive versions of a service interface in order to detect changes that cause clients using an earlier version not to interact properly with a later version. We focus on behavioural incompatibilities and adopt the notion of simulation as a basis for determining if a new version of a service is behaviourally compatible with a previous one. Unlike prior work, our technique does not simply check if the new version of the service simulates the previous one. Instead, in the case of incompatible versions, the technique provides detailed diagnostics, including a list of incompatibilities and specific states in which these incompatibilities occur. The technique has been implemented in a tool that visually pinpoints a set of changes that cause one behavioural interface not to simulate another one.

Enabling Interoperability for SOA-Based SaaS Applications Using Continuous Computational Language

Today there still exist many interoperability issues concerning the communications between Web Services (WSs). Most of the existing approaches can only solve either the structural or behavioural interoperability conflicts between SOAP messages that are exchanged by heterogeneous WSs. Furthermore, they are usually ad-hoc approaches that are too costly and not reusable elsewhere. In this demonstration we present a novel prototype tool that allows encoding adaptation rules in the Continuous Query Language (CQL) for SOAP messages on both structural and behavioural levels, and then deploying and managing the rules on a rule engine in a predictable and repeatable manner.

ArchiMed : canevas multimédiateur pour la réconciliation de conversations entre services web

web service-based technology is widely accepted as the infrastructure which enables applications interoperability. In this setting, message exchanges form the basics of interac- tions between applications wrapped as web services. The set of messages exchanged between two services describes a conversation. The study reported in this text aims at reconciliating conversations which fail due to an evolution of the provided interface which is thus no longer compatible with the one required by clients. To address this issue, we propose a framework which, relying on successive versions of the interface provided by the service, generates media- tors. When a client starts a conversation, the framework selects one suitable mediator which is responsible for seamlessly resolving incompatible conversations. This article presents a model and its implementation for detection and resolution of incompatibilities.