José Antonio Martín

ITACA: An integrated toolbox for the automatic composition and adaptation of Web services

Adaptation is of utmost importance in systems developed by assembling reusable software services accessed through their public interfaces. This process aims at solving, as automatically as possible, mismatch cases which may be given at the different interoperability levels among interfaces by synthesizing a mediating adaptor. In this paper, we present a toolbox that fully supports the adaptation process, including: (i) different methods to construct adaptation contracts involving several services; (ii) simulation and verification techniques which help to identify and correct erroneous behaviours or deadlocking executions; and (iii) techniques for the generation of centralized or distributed adaptor protocols based on the aforementioned contracts. Our toolbox relates our models with implementation platforms, starting with the automatic extraction of behavioural models from existing interface descriptions, until the final adaptor implementation is generated for the target platform.

Automatic Generation of Adaptation Contracts

Software development based on the composition of black-box software like Web Services and Software Components is impeded by incompatibilities in their interfaces. Software adaptation has emerged as a solution to these incompatibilities by using processes in-the-middle (called adapters) that allow the correct interaction between the services. There are several approaches that focused on the automated generation of adapters guided by adaptation contracts which specify how the incompatibilities can be resolved. However, the generation of these contracts is not automated and most existing approaches require these contracts to be specified by hand, which obliges the designer to know all the service details. In this paper, we propose an approach to automatically generate adaptation contracts from the behavioral description of the services. These contracts overcome incompatibilities at signature and behavioral levels. Finally we present our prototype tool that accepts as input the service behaviors written in abstract BPEL and generates adaptation contracts using a combination of an A* algorithm and an expert system.

Contracts for security adaptation

Abstract Security is considered to be one of the main challenges as regards the widespread application of Service Oriented Architectures across organisations. WS-Security, and its successive extensions, have emerged to fulfil this need, but these approaches hinder the loose-coupling among services, therefore constraining their reusability and replaceability. Software adaptation is a sound solution to overcome the incompatibilities in interface, behaviour and security constraints among stateful services. However, programming adaptors from scratch is a tedious and error-prone task where special care must be given to concurrency and security issues. In this work, we propose to use security adaptation contracts that allow us to express and adapt the security requirements of the services and their orchestration. Given a security adaptation contract and the behavioural description of the services (such as BPEL processes or Windows Workflows), we can generate the protocol of the orchestrator that complies with the security requirements (confidentiality, integrity and authenticity), while overcoming incompatibilities at the signature, behaviour and security QoS levels. The formalisation behind security adaptation contracts has other applications such as security policy negotiation and automatic security protocol verification.

Synthesis of secure adaptors

Abstract Security is considered one of the main challenges for software oriented architectures (SOA). For this reason, several standards have been developed around WS-Security. However, these security standards usually hinder interoperability, one of the main pillars of Web service technologies. Software adaptation is a sound solution where an adaptor is deployed in the middle of the communication to overcome signature, behavioural and QoS incompatibilities between services. This is particularly important when dealing with stateful services (such as Windows Workflows or WS-BPEL processes) where any mismatch in the sequence of messages might lead the orchestration to a deadlock situation. We proposed security adaptation contracts as concise and versatile specifications of how such incompatibilities must be solved. Nonetheless, synthesising an adaptor compliant with a given contract is not an easy task where concurrency issues must be kept in mind and security attacks must be analysed and prevented. In this paper, we present an adaptor synthesis, verification and refinement process based on security adaptation contracts which succeeds in overcoming incompatibilities among services and prevents secrecy attacks. We extended the ITACA toolbox for synthesis and deadlock analysis and we integrated it with a variant of CCS, called Crypto-CCS, to verify and refine adaptors based on partial model checking and logical satisfiability techniques.

Semi-Automatic Specification of Behavioural Service Adaptation Contracts

An adaptation contract describes composition constraints and adaptation requirements among several services which were not initially built to interoperate with each other. The manual writing of this specification by a designer is a difficult and error-prone task, especially when services are reused taking their behavioural descriptions into account. In this paper, we present a semi-automatic approach to build adaptation contracts. To this purpose, we propose an adaptation contract design process supported by an interactive environment based on a graphical notation, and an engine capable of automatically generating contracts without any human intervention. We also present an experimental study that we carried out using the tool support that we implemented in order to evaluate our approach.

A tool for the synthesis of cryptographic orchestrators

Security is one of the main challenges of service oriented computing. Services need to be loosely coupled, easily accessible and yet provide tight security guarantees enforced by cryptographic protocols. In this paper, we address how to automatically synthesize an orchestrator process able to guarantee the secure composition of electronic services, supporting different communication and cryptographic protocols. We present a theoretical model based on process algebra, partial model checking and logical satisfiability, plus an automated tool implementing the proposed theory.

Automated Synthesis and Ranking of Secure BPMN Orchestrators

We describe a formal methodology for the automatic synthesis of a secure orchestrator for a set of BPMN processes. The synthesized orchestrator is able to guarantee that all the processes that are started reach their end, and the resulting orchestrator process is secure, that is, it does not allow discloure of certain secret messages. In this work we present an implementation of a forth and back translation from BPMN to crypto-CCS, in such a way to exploit the PaMoChSA tool for synthesizing orchestrators. Furthermore, we study the problem of ranking orchestrators based on quantitative valuations of a process, and on the temporal evolution of such valuations and their security, as a function of the knowledge of the attacker.

Learning from Failures: A Lightweight Approach to Run-Time Behavioural Adaptation

Software integration needs to face signature and behaviour incompatibilities that unavoidably arise when composing services developed by different parties. While many of such incompatibilities can be solved by applying existing software adaptation techniques, these are computationally expensive and require to know beforehand the behaviour of the services to be integrated. In this paper we present a lightweight approach to dynamic service adaptation which does not require any previous knowledge on the behaviour of the services to be integrated. The approach itself is adaptive in the sense that an initial (possibly the most liberal) adaptor behaviour is progressively refined by learning from failures that possibly occur during service interaction.

A case study in model-based adaptation of web services

Developing systems through the composition of reusable software services is not straightforward in most situations since different kinds of mismatch may occur among their public interfaces. Service adaptation plays a key role in the development of such systems by solving, as automatically as possible, mismatch cases at the different interoperability levels among interfaces by synthesizing a mediating adaptor between services. In this paper, we show the application of model-based adaptation techniques for the construction of service-based systems on a case study. We describe each step of the adaptation process, starting with the automatic extraction of behavioural models from existing interface descriptions, until the final adaptor implementation is generated for the target platform.

Dinapter: Automatic Adapter Specification for Software Composition

Enterprise systems rely heavily on compositional software like Software Components or Web Services. The composition of this software allows software reusability, greater productivity and reduced costs. However, these components are black-boxes and their direct reuse is prevented by incompatibilities between their interfaces. There are several approaches focused on Software Adaptation which are capable of solving incompatibilities at signature and behavioral levels, but these approaches require abstract specifications which specify how the incompatibilities can be resolved. The generation of these specifications is an open issue and specifications are normally handmade, which forces the designer to understand the subtleties of the components. In this paper we present Dinapter, a tool that automatically generates specifications being given the component behavioral descriptions written in abstract BPEL. Dinapter complements the aforementioned approaches and allows the automatic adaptation of compositional software.