Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: New Results

Adaptive Middleware

Participants : Damien Fournier, Rémi Mélisson, Philippe Merle, Russel Nzekwa, Lionel Seinturier, Daniel Romero, Romain Rouvoy.

Three results are worth mentioning around Adaptive Middleware. They deal with the FraSCAti platform, autonomic feedback control loops and complex event processing.

Concerning the FraSCAti middleware platform for adaptive SOA, in [25] we describe some new results which have been incorporated in the platform to support the autonomic discovery of services. We have built a demonstration scenario which highlights the adaptation capabilities of FraSCAti, and two new forms of remote bindings: resource-oriented and ubiquitous. This demonstration scenario is architectured around a case study in the domain of an ubiquitous home environment with devices such as set top boxes and smartphones.

In the context of the ANR SALTY project we worked on the specification of a DSL named CORONA and dedicated to the engineering of control loops in autonomic systems. The purpose of CORONA is twofold: First, it aims to reduce the cost of building autonomic administrative software. Secondly, it aims to enforce visibility of control loops in such software. Preliminary results of this work have been published in [27] .

For Complex Event Processing (CEP), we have proposed in [21] and  [22] a plug-in based framework, called CEVICHE, that allows the integration of different Complex Event Processing (CEP) engines, to create context-adaptive business processes, which are described with the Standard Business Process Language (SBPL) that we have proposed. SBPL is an extension of BPEL, which allows the user to include, in the business process definitions, the adaptation points, conditions and alternative processes to create adaptable business processes dynamically. We used two different approaches, AOP and SCA, to demonstrate the feasibility of business processes adaptability at runtime. This work is partially funded by the ICOM project.


Logo Inria